How to tell whether your child actually understands maths, or is just pattern-matching

The distinction

Pattern-matching is not cheating and it is not laziness. It is an intelligent response to how maths is usually taught and tested. A child sees thirty questions that all yield to the same move, notices this, and stops attending to the questions. This is efficient. It is also, for several years, completely invisible, because it produces correct answers.

A child who understands is doing something different: they are reading the question, forming a picture of what it is asking, and choosing a method because it fits. When the question is unfamiliar they slow down. When it is unfamiliar for a pattern-matcher, they guess at which learned shape it most resembles — and that guess is often wrong in a way that looks like carelessness.

The reason this matters is that the two strategies have completely different failure curves. Pattern-matching is more reliable than understanding in Class 6, because the patterns repeat and the questions are drawn from a small set. By Class 9, when questions start combining ideas, the set stops being small and the strategy collapses — apparently suddenly, though nothing sudden happened.

Three children

All examples below are from our own sessions, anonymised, and published with the family's consent.

Five questions you can ask tonight

None of these are hard questions. That is the point — difficulty is not what separates the two children. Ask them on paper, and ask for reasoning rather than answers.

1. "Is the answer going to be bigger or smaller than what we started with? Don't work it out." Understanding predicts. Pattern-matching cannot answer this at all, because the procedure has not been run yet.
2. "Why does that work?" Asked once, gently, about something they have just got right. The answer "because that's the rule" is information, not insolence.
3. "Can you show me that with a drawing?" A procedure does not have a picture. An idea usually does, even a clumsy one.
4. "I'm going to change one number. Does your method still work?" Change something that breaks the shape — make a number negative, or make it not divide evenly. Understanding adapts. Pattern-matching applies the same steps and arrives somewhere strange without noticing.
5. "Here's a wrong answer someone else gave. Why do you think they thought that?" The hardest of the five and the most revealing. Diagnosing someone else's error requires a model of the idea, not a procedure.

What you are listening for across all five is not correctness. It is whether there is a reason underneath, and whether the reason is about quantities or about steps.

Why school assessment misses it

Not through negligence. A written test has to be markable at scale, which means it has to have right answers, which means it rewards the production of right answers by any route. A child who reasons and a child who matches produce the same mark, and the mark is the only thing that travels home.

Worse, the two get separated by time. The gap opens in Class 5 or 6 and the consequence lands in Class 8 or 9, by which point everyone — parents, teachers, and the child — is looking at the topic in front of them. Nobody goes back three years, because nothing three years ago looked like a problem. Every mark was fine.

This is the whole reason we work from wrong answers rather than scores. A right answer is compatible with both strategies and therefore tells you very little. A wrong answer is specific: it shows which idea the child was actually using, and it is almost always a reasonable idea applied where it does not belong.

What to do about it

If the five questions suggest your child is pattern-matching, the instinct is to add practice. Resist it. More practice on the same shapes strengthens exactly the strategy you are trying to replace.

Instead, find the specific earlier idea that is missing. It is usually one thing, two or three classes back, and fixing it is a matter of weeks rather than a matter of starting again. The topic pages on this site are built for this: each one lists what a topic actually requires and gives you a five-minute check for each prerequisite.

Common places to start, depending on where the difficulty showed up:

• Equivalent fractions — if fractions of any kind are the problem
• Dividing fractions — the topic where the gap between procedure and meaning is widest
• Negative numbers on a number line — if sign errors persist into algebra
• What a variable actually means — if algebra arrived and never landed

And one thing that costs nothing: when your child gets something right, ask why about one question in five. Not as a test. Just as interest. Children who are asked for reasons start producing them.