Most children losing marks in PSLE maths can do the arithmetic perfectly well. What they cannot do is turn a paragraph of text into something solvable — decide what is being asked, represent the relationships, and choose an approach. That is a different skill from computation, and drilling more sums does not build it.
Knowing which of the two your child is missing changes everything about what to do next.
The two very different problems
Computation. Can the child carry out the operations reliably — fractions, ratio, percentage, decimals, area and volume — without errors?
Problem representation. Given a multi-step word problem, can the child work out what is happening, represent it, and see a route to the answer?
A child weak in the first needs practice. A child weak in the second needs to be taught a method for attacking unfamiliar problems, and more computation practice will not touch it. The second is far more common at P5–P6, and it is routinely misdiagnosed as “weak in maths”, which leads families to buy exactly the wrong intervention.
The distinction is easy to test. Take a word problem your child got wrong and read it aloud, then ask them to explain what is going on in the question — not to solve it. If they can explain the situation and still could not start, the gap is method. If they cannot explain the situation, the gap is comprehension of the problem itself. If they explained it, chose a correct route, and still got it wrong, the gap is execution.
What upper-primary maths actually demands
The step up at P5–P6 is in the number of moves a question requires and in how little the question signals about method.
Multi-step reasoning. A question that requires three linked operations, where an error at any step fails the whole thing, and where none of the steps are labelled.
Representation. The model method — drawing bar models to represent quantities and their relationships — is the backbone of Singapore primary maths problem-solving. A child fluent with it can make a great many otherwise opaque problems visible. A child who half-learned it and abandoned it is working much harder than necessary.
Heuristics. Working backwards, guess-and-check, looking for a pattern, simplifying the problem. These are teachable strategies for starting when the route is not obvious, and a child with no starting strategy simply stalls on unfamiliar questions.
Transfer. Recognising that a question dressed in a new context is structurally one they have seen before. This is the single hardest thing to teach and the thing that separates outcomes most.
Sorting where the marks went
Take a recent paper and sort every lost mark into four piles. The proportions decide what to buy.
Could not start. No route came to mind. Method and heuristics gap.
Started wrong. Misread the question or represented it incorrectly. Comprehension and representation gap.
Right method, wrong answer. Arithmetic slips, sign errors, dropped units. Execution.
Ran out of time. Often downstream of the first pile — time spent stalled on questions with no starting strategy.
A child whose losses sit mostly in piles three and four will gain little from more teaching and quite a lot from timed practice with strict checking. A child stuck in pile one needs method taught explicitly, and that is where tuition earns its cost.
What actually helps, by pile
For “could not start”: explicit teaching of the model method until it is genuinely fluent, plus a small set of heuristics practised on unfamiliar problems. Fluency matters — a half-remembered technique is worse than none, because the child abandons it mid-problem.
For “started wrong”: slow, deliberate reading of the question, underlining what is asked, and stating the question in their own words before touching a pencil. Much of this is a habit, not a skill.
For “right method, wrong answer”: checking habits. Re-reading the question at the end and asking whether the answer is plausible catches a large share of these. Units and labels matter.
For “ran out of time”: timed practice under real conditions once the method gap is closed. Doing this before the method gap closes just produces faster stalling.
On the pressure
PSLE is a high-stakes examination and the anxiety around it is real. Two things are worth holding onto.
Adding hours has a ceiling, and it can go negative. A tired child with no time to consolidate learns less per hour. Families sometimes add a third weekly session in the final months and see results fall, which is not a mystery.
A child who has decided they are bad at maths is working against that belief before they start. That is a real obstacle and more sessions do not address it — being able to solve a problem they previously could not does. Small, visible wins on problem types they used to stall on shift more than volume does.
Helping without knowing the method yourself
Many parents were taught differently and worry that stepping in will confuse the child. You do not need the method to be useful — the most valuable things a parent does here require no maths at all.
Ask them to explain the question, not solve it. “What’s happening in this problem?” A child who can describe the situation and still cannot start has a method gap. A child who cannot describe it has a comprehension gap. That single question separates the two, and you need no arithmetic to ask it.
Ask them to show you the drawing. If a bar model was attempted, look at whether it represents the relationships or is decorative. You do not need to check the maths — you need to see whether the child used a representation at all.
Ask what they tried first, and why. A child with a starting strategy will answer. A child without one says “I didn’t know what to do”, which is the diagnosis.
Resist teaching your own method. A parent’s alternative approach can work, and it can also leave the child holding two half-methods. If you want to help directly, ask questions rather than supply routes.
Let them explain a question they got right. Confidence is part of the work, and articulating a correct method reinforces it more than correcting a wrong one.
What to ask a prospective tutor
- “After a few sessions, can you tell me whether the gap is method, representation or execution?” A tutor who answers this specifically is diagnosing. One who says “we’ll strengthen her foundations” may not be.
- “How do you teach a child to start an unfamiliar problem?” Listen for something concrete — model drawing, a heuristic sequence — rather than “lots of practice”.
- “Is work marked between sessions?” Marking is where the diagnosis actually happens.
Frequently asked questions
My child knows the topics but still loses marks. Why? Almost always representation or execution rather than knowledge. Knowing fractions and being able to see that a word problem is a fractions problem are different abilities, and PSLE tests the second.
Is the model method still worth teaching if my child dislikes it? Fluency is what matters, not preference. A child with a reliable alternative route can use it, but a child who abandons the model method without replacing it usually has no way to start hard problems.
How early should we start preparing? P5 is where the reasoning demand steps up, so a gap that appears then is worth addressing then. Starting intensive preparation only in P6 leaves the method work compressed into the period when timed practice should be happening.
Should we do past papers repeatedly? Useful once the method gap is closed — for timing, stamina and transfer. Before that, past papers mainly confirm the child cannot start, which you already know.
How many hours a week is reasonable? Enough to close the identified gap, with time left to consolidate. If the child has no unstructured time at all, additional hours are likely reducing total progress rather than adding to it.
—