She finished every worksheet. She never complained. And she had no idea what she was doing. Imagine this. A 6th grader — bright kid, well-liked, no history of academic trouble — sits down to do her weekly review. She’s been going to a well-known tutoring center for nearly two years. Her parents are proud. Her worksheets come home looking great. But today, the review question is structured a little differently. Not harder. Just different. The numbers are the same. The concept is the same. The presentation is not. She stares at it for a long time. And then she puts her pencil down. “I don’t know how to do this one,” she says. Her mother was floored. Two years. Two hundred dollars a month. A child who can complete sixty problems in under fifteen minutes. And when the environment shifted by ten percent, she had nothing to reach for. | | “She wasn’t learning math. She was mirroring the environment she’d been conditioned to expect.” |
I want to name this plainly, because I think it deserves to be named: this is what Kumon produces. I say this not to be harsh, but because I am genuinely bothered by it. The Kumon model is built on volume and repetition — hundreds of near-identical problems, timed, drilled, repeated until the process is automatic. And for a certain kind of parent, watching a child blaze through that stack of papers feels like progress. It isn’t. What it produces is what researchers call procedural conditioning — the child has memorized the steps of a narrow pathway. Remove the scaffolding, vary the problem format, ask them to explain why rather than how, and the whole structure collapses. Worse, many of these children start to associate math with tedium and anxiety. They learn to dread it because it was never made meaningful. THE RESEARCH Studies on math achievement distinguish between procedural fluency — executing steps correctly — and conceptual understanding — knowing why those steps work. Children with only procedural fluency perform well on familiar problems but struggle significantly on novel ones. Conceptual understanding transfers. Procedure alone does not. |
At SingMath, we don’t sell speed. We don’t promise tearless homework nights. What we promise is something far more durable: internal architecture. When one of our students in our small group of six encounters a problem they’ve never seen before, they have tools — real mental tools, not memorized sequences — to reach for. They can draw a bar model. They can sketch a number bond. They can reason through it. The silence when a child wrestles with a hard problem — the genuine wrestling, not the rote recall — that silence is not failure. That silence is the sound of architecture being built. We trust our students enough to let them sit in it. And that trust is what changes everything. On the blog, I’ve gone much deeper into how this plays out across different grade levels — and specifically what a parent can look for to know whether their child has conceptual ownership versus procedural mimicry. If the story of that 6th grader made you pause, I think you’ll find it valuable. | 
