You do not need to know the name of the problem before you can begin fixing it.
A learner can say, “I am bad at algebra,” “I keep losing marks,” “I understand in tuition but not in tests,” or simply, “I don’t know what went wrong.” Those are useful signals, but they are not yet diagnoses.
This page is a routing page for Bukit Timah Tutor. Its job is not to replace the Mathematics Diagnosis, the Mathematics HELP Runtime, the How Mathematics Works machine, or the Singapore Mathematics Hub. Its job is simpler: help a learner, parent or tutor decide where to look next.
Start with the state, not the score
A score is an outcome. A mathematics state is a working description of what is currently happening inside the learning process.
Two students can both score 45 marks and need completely different next actions. One may have a missing prerequisite from two years earlier. Another may know the mathematics but misread task language under time pressure. A third may reproduce familiar procedures but fail when the question changes form.
So BTT begins with a more useful question:
What is the earliest thing that is unstable enough to explain the errors we can actually see?
That is the beginning of diagnosis. It keeps us from treating every low mark with more worksheets and every high mark as proof that the structure is secure.
The eight common mathematics states
These are not labels for children. A learner can move between them, occupy more than one at once, and leave them after repair. They are simply useful starting states for deciding what evidence to collect next.
1. The prerequisite is missing
The current topic looks difficult because an earlier object is not stable. Fractions may disturb algebra. Ratio may disturb percentage. Signed numbers may disturb expansion. Equation solving may disturb coordinate geometry. The visible failure appears late; the useful repair point may be much earlier.
Useful next move: do not reteach the whole chapter. Use the Mathematics Diagnosis to search backward for the earliest weak dependency.
2. The representation is fragile
The learner may know a rule but not see what the symbols, graph, table, diagram or equation are representing. This often creates the feeling that mathematics changes from question to question, even when the underlying relationship has not changed.
Useful next move: return to the mathematics machine in How Mathematics Works: represent, relate, operate, generalise, model, solve and verify.
3. The procedure works only in familiar form
The learner can complete rehearsed exercises but stalls when values, wording, layout or context change. This is not necessarily a lack of effort. It can be a sign that the procedure has been memorised more strongly than the relationship that makes the procedure valid.
Useful next move: use changed-form questions, mixed practice and independent attempts before assuming the topic is secure. The BTT Mathematical Lab exists for this observe–probe–repair–validate–release cycle.
4. The transfer is breaking
The learner understands the idea in one chapter or one teaching context but does not recognise when it should be used elsewhere. A proportion idea may be known in ratio but missed in similarity. Gradient may be known in coordinate geometry but not recognised as rate of change later. Algebraic structure may be understood in isolation but not used inside a modelling problem.
Useful next move: test the same mathematical relationship across different representations and contexts. Transfer is evidence that the learner owns more than the surface form.
5. The task language is interfering
The mathematics may be available, but the learner misreads a condition, overlooks a restriction, confuses “hence” with a fresh start, answers the wrong quantity, or cannot convert a verbal condition into mathematical structure.
Useful next move: separate comprehension from calculation. Ask the learner to identify what is known, what is constrained, what is required, and what representation would make those relationships visible before calculation begins.
6. The mathematics is available, but examination execution is unstable
The learner can solve the question at home but loses accuracy, pacing, checking discipline or answer-form control under examination conditions. That is a different problem from not knowing the mathematics.
Useful next move: route toward BTT’s examination layer rather than restarting the entire topic. Use the mathematics hub and examination-specific guides to work on command words, answer forms, pacing, calculator discipline, verification and paper architecture.
7. Too much help is carrying the learner
The learner appears successful while a tutor, parent, worked example, solution key or AI system is continuously supplying the next step. When support disappears, the performance disappears with it.
Useful next move: use the Mathematics HELP Runtime. HELP should be the minimum justified intervention: enough to restore productive movement, then gradually removed so the learner must carry the mathematics again.
8. The learner is stable and ready to stretch
Not every diagnostic route begins with failure. Sometimes the evidence says the current work is secure. The correct action is then not endless repetition. It may be deeper connections, unfamiliar problems, stronger proof, modelling, more demanding transfer, Additional Mathematics, JC mathematics or wider mathematical exploration.
Useful next move: enter the Singapore Mathematics Hub and move outward only after the current layer is demonstrably stable.
A 90-second routing check
Use the first statement that sounds most like the present situation. Do not try to make every statement fit.
- “I never really understood the earlier idea.” → search prerequisites.
- “I know the formula but I don’t see what the question is doing.” → test representation and relationships.
- “I can do the worksheet but not a different-looking question.” → test transfer.
- “I knew it yesterday but cannot retrieve it now.” → test retention and practice spacing.
- “I keep doing the wrong thing even though I know the topic.” → inspect task language and decision points.
- “I can do it slowly but not in an exam.” → inspect execution and examination craft.
- “I can do it only when someone helps me.” → reduce support through HELP.
- “This is easy now; what should I do next?” → stretch and connect.
If none of these is clear, collect better evidence. One marked paper, one school worksheet, one recent test, or three carefully chosen questions can be more useful than a vague description of “weak Mathematics”.
What counts as useful evidence?
BTT diagnosis should work from observable mathematics rather than personality labels. Useful evidence can include:
- the exact line where a correct solution first becomes incorrect;
- a repeated error across different questions;
- a representation the learner avoids or misreads;
- a prerequisite question that changes the likely diagnosis;
- a correct answer reached by an unstable method;
- a problem solved with help but not independently;
- a familiar question solved correctly and a transfer question solved incorrectly;
- an examination error that disappears when time pressure is removed.
The goal is not to collect as many mistakes as possible. The goal is to find the smallest amount of evidence that meaningfully changes what we should do next.
Do not repair everything at once
A learner can have many visible errors. That does not mean every error deserves a separate intervention. Sometimes five later mistakes share one earlier cause. Fixing the cause can collapse several downstream problems at once.
This is why BTT distinguishes between a weak mathematical object and a weak connection between objects. A student may know fractions and know algebra but fail at the connection between them. A student may know a graph and know an equation but not reliably translate one into the other. The useful repair is not always another chapter of notes; sometimes it is the missing edge.
When the evidence is uncertain, use a discriminating question. A good diagnostic question is not merely “hard”. It is chosen because different answers point toward different explanations.
The BTT route after a state is identified
Once a likely state is identified, the learner moves into the owner that can do the real work.
| Need | Owner | Job |
|---|---|---|
| Where do I begin? | Start Here at BTT | Route by learner, stage and immediate need. |
| What is the mathematical problem? | Mathematics Diagnosis | Find the earliest useful weak link from evidence. |
| How much help should be given? | Mathematics HELP Runtime | Provide the minimum justified help, then fade it. |
| How does the subject itself work? | How Mathematics Works | Represent, relate, operate, generalise, model, solve and verify. |
| Where are the learning guides? | Singapore Mathematics Hub | Own the wider Mathematics library and stage routes. |
| How do we test repair? | BTT Mathematical Lab | Observe, probe, repair, validate and release. |
For parents: ask for the next useful decision
A useful parent conversation does not need to begin with “How many marks can my child gain?” A more diagnostic sequence is:
- What evidence are we using?
- Where does the work first become unstable?
- Is that a missing object, a weak relationship, a transfer problem or an execution problem?
- What is the smallest repair worth testing?
- What changed question will show whether the repair held?
- Can the learner now do it with less help?
This keeps the conversation close to observable learning. It also makes progress easier to verify because the intervention has a predicted result.
For learners: “I don’t know” is enough to begin
You do not need to arrive with a perfect explanation of your weakness. Bring the work. Show the point where you became unsure. Show what you tried. Show what happened when the question changed.
A good diagnosis becomes more precise as evidence improves. It should be allowed to change when new evidence contradicts the first explanation.
The end point is not permanent support
BTT’s mathematics system should eventually make itself less necessary for the task that has been learned. The learner should increasingly be able to represent the problem, choose a route, monitor errors, verify the result and decide when something genuinely does not make sense.
That is why this page routes toward diagnosis and HELP rather than becoming another large collection of explanations. The purpose of the route is to return the mathematics to the learner.
Choose the next door
- Start Here at Bukit Timah Tutor — if you need the broadest route.
- How Mathematics Diagnosis Works — if the first weak link is unclear.
- Mathematics HELP Runtime — if you need the right amount of intervention.
- How Mathematics Works — if you need to understand the machine behind the subject.
- Singapore Mathematics Hub — if you know the stage or topic and want the learning estate.
BTT Runtime principle: identify the state, collect enough evidence, repair the earliest useful weakness, test the prediction, reduce help, and return control to the learner.