Your Child Is Not Learning Math

She’s Just Mirroring It.

This article expands on a private letter sent to our SingMath parents. Read the original letter here.

There is a question every parent of a math student should be able to answer, and almost none of them can: Is my child learning math, or are they just executing a familiar sequence? These are not the same thing. They look identical from the outside — the worksheets come home complete, the grades are fine, the child doesn’t complain — but they diverge catastrophically the moment the environment changes even slightly.

This distinction is the heart of what we do at SingMath Tutoring. And it is exactly what high-volume drill programs like Kumon consistently fail to address — and in doing so, actively work against the children they claim to help.

The Difference Between Procedural Fluency and Conceptual Ownership

Mathematics education researchers have long distinguished between two kinds of competence. Procedural fluency means a student can execute a learned sequence of steps correctly and quickly. Conceptual understanding means a student grasps why those steps work — what mathematical truth underlies them — and can therefore adapt when the steps look different or need to be applied in an unfamiliar context.

A child with strong procedural fluency but weak conceptual understanding is, in the most precise sense, performing mimicry. They have observed a pattern, internalized it as a reflex, and can reproduce it reliably — as long as the stimulus matches the conditioned environment. Change the stimulus, and the reflex has nothing to fire at.

This is not a small problem. This is the difference between a child who can function mathematically as the material scales in complexity, and one who hits a wall — often around 5th or 6th grade — and cannot understand why. The parent cannot understand why either, because the worksheets looked so good.

Drill builds a reflex. A bar model builds a mind.

Why Kumon Produces Brittle Learners

I want to be direct here, because I think the math education conversation in this country is too polite about this. Kumon is the most widely used supplemental math program in the United States. It is built on a single pedagogical premise: that mathematical competence is built through the repetition of slightly progressing problems, drilled until the execution becomes automatic.

This premise is not entirely wrong. Automaticity in basic arithmetic — knowing that 7 × 8 = 56 without having to reconstruct it — does free up cognitive bandwidth for higher-level reasoning. The problem is that Kumon stops there. The automatic execution of steps becomes the product. There is no exploration of why the steps work. There is no problem variation designed to test transfer. There is no architectural thinking — no visual modeling, no decomposition into parts, no building of mental structures that survive contact with the unfamiliar.

What Kumon produces is what learning scientists call context-dependent knowledge: competence that is tightly bound to the exact stimuli under which it was acquired. Many Kumon students also develop something arguably worse: a deep association between mathematics and tedium. They learn that math is a chore to be endured, a pile of identical problems to be dispatched as fast as possible. That association does not disappear when they graduate from worksheets. It follows them into middle school, high school, and college — and it is a primary driver of the STEM attrition crisis that costs this country dearly every year.

What the Singapore Method Actually Does to a Brain

The Singapore Mathematics framework — the methodological backbone of what we teach — was engineered from the ground up to build transferable, conceptual understanding. It does this through three interconnected principles.

The first is the CPA (Concrete-Pictorial-Abstract) progression. A child doesn’t begin with abstract numerals and operations. They begin with concrete objects — real quantities, manipulated and grouped — and only after genuine contact with the concept do they move to pictorial representation (like a bar model), and only after that to the abstract symbolic notation. This sequence ensures that when a child writes “48 ÷ 6 = 8,” they have already felt what division means in two prior registers. The abstract symbol is the last layer, not the first.

The second principle is bar modeling — a visual problem-solving tool that renders abstract relationships as spatial diagrams. A bar model — for example, a rectangle labeled 63 total, split cleanly into sections of 40 and 23 — forces a student to construct the relationship between quantities before calculating anything. It is a thinking tool, not a computation tool. Students who learn to model problems visually develop an enormously flexible mental habit: when they encounter a new problem, their first instinct is to represent the structure, not to reach for a memorized procedure.

The moment a bar model clicks is the moment math stops being a mystery.

The third principle is problem variation. In our small groups — capped at six students — we regularly present problems whose surface features change while the underlying structure remains constant. We might ask the same quantity relationship using apples one day and running distances the next. This is not random; it is a deliberate pedagogical technique grounded in cognitive science. Varied practice forces the brain to extract the abstract structure of the problem rather than pattern-matching to a specific surface form. It is precisely what builds the kind of knowledge that transfers — the kind that survives an unfamiliar worksheet, a standardized test, a college exam, a career in STEM.

The Struggle Is the Point

Most tutoring services sell relief. They promise fewer tears, smoother homework nights, better grades with less friction. I understand the appeal. Parenting is exhausting, and watching a child struggle with a math problem at 9pm is painful.

But here is the truth: the struggle is not a bug in the process. It is the process. When a child in one of our small groups pauses over a problem they haven’t seen before — when their cursor hovers on the screen, when they go quiet, when you can almost see them reaching inward — that pause is where architecture gets built. We call it productive struggle. It is the precise moment when a child is forced to reason rather than recall. And it is the only moment when conceptual understanding is actually created.

We trust our students enough to let them sit in that silence. We do not rescue them the moment the path looks uncertain. We ask questions. We wait. We watch them find their way. And when they do — when the bar model comes together, when the number bond resolves, when the student says “oh, I see it now” — that is not just a solved problem. That is a child who has just built a piece of durable mathematical architecture that no future test, no unfamiliar problem format, and no change in environment can take away from them.

How to Know Which Kind of Learning Your Child Has

If you want to know whether your child has conceptual ownership or procedural mimicry, try this: take a problem they recently solved correctly, change a surface feature — the numbers, the context, the orientation of the question — and ask them to solve it without you helping. Then ask them to explain what they did and why.

A child with genuine understanding will slow down, think, and work through it. They may not be fast. They may make a false start. But they will be reasoning. A child operating on procedural mimicry will either get it quickly because the surface pattern still matches close enough, or will go blank because the trigger they were conditioned to respond to isn’t present.

You should not have to be a math educator to run that test. And if the result troubles you, I want you to know that conceptual gaps — even significant ones — are absolutely repairable with the right approach. That is exactly what our free consultation and student evaluation is designed to surface and address.

We are not in the business of making math feel easy. We are in the business of making children capable. That is a completely different thing, and it lasts considerably longer.

With genuine love for what we do,

Leeling

Director of Education & Academic Standards
SingMath Tutoring, home of Singapore Math Classes & Camps