She Passes the Tests. Is She Actually Learning?

This is an awkward worry to admit to, because on paper there's nothing wrong. The reports are fine. The grades are fine. Nobody has called you in.

And yet something nags. She can pass the test on Friday and be visibly blank about the same material a month later. He can do the exercises in the chapter and then not recognise the identical idea when it appears in a different shape.

You're not imagining it, and you're not being ungrateful about a child who's doing well. You've noticed a real thing that assessment is quite bad at catching.

Why can a child pass a test without understanding?

Because a test measures what's retrievable, in that format, on that day. Several things can produce a good mark, and only one of them is understanding.

Recency. Material revised the night before is sitting in easily accessible memory. That's a genuine achievement of short-term effort — it just isn't the same as knowing something.

Pattern-matching. Enormous numbers of school questions are recognisable by shape. A child can learn that "these numbers and this phrasing" means "do the thing with the two steps" without ever knowing why the two steps work. This gets you a long way, and it collapses the moment the question is phrased unusually.

Format familiarity. Practising past papers trains you at past papers. Some of that is real transferable skill; some of it is knowing where the marks are hidden.

Narrow sampling. A test asks a handful of questions from a large topic. A child can have three solid islands of knowledge in a sea of nothing and, with reasonable luck, be sampled entirely on the islands.

None of this means your child is cheating or coasting. These are rational strategies for succeeding at the thing they're actually being asked to do. The system rewards them, so they use them.

What's the difference between a grade and mastery?

A grade tells you where your child placed relative to other children, on one occasion, in one format.

Mastery tells you what your child can actually do — reliably, in a new context, without prompting.

These are not the same thing, and they come apart constantly. A child can be top of a weak class and have shaky foundations. A child can be middling in a strong class and understand the material perfectly well.

Worse, grades are comparative, which makes them nearly useless as a signal about an individual. "Above average" tells you about the average. It tells you very little about the child.

A grade tells you where your child sat in a ranking. Mastery tells you what they can actually do.

How can I check what my child really understands?

You don't need to test them. You need to make them use the knowledge somewhere it isn't already loaded.

Ask them to teach it to you. The single best check there is. Explaining forces you to sequence ideas and expose the joins. A child who's pattern-matched will produce a description of the procedure — "you take this number and put it here" — with no account of why. A child who understands can say what would happen if you changed something.

Ask "why" one level deeper than the lesson went. Not to catch them out; to see whether there's structure underneath. "Why does that rule work?" If the answer is "because that's the rule," you've found the edge.

Change the surface. Same concept, different clothing. Fractions as pizza, then as a measuring jug, then as a bar chart. Understanding survives a costume change; memorisation doesn't.

Wait, then ask. Come back to something three or four weeks later. Anything that survives that gap without revision is genuinely learned. This one is uncomfortable, because a lot won't survive — that's normal, and it's the most honest signal you'll get.

Ask them to find the mistake. Give them a worked answer with an error in it. Spotting a flaw requires actually knowing how the thing should go.

Should I be worried if the answers are bad?

Less than you'd think, and it depends what kind of bad.

Forgetting is not failure — it's the default behaviour of human memory. Almost everything fades without retrieval. That a child can't recall October's topic in March is expected, not alarming.

What matters is whether it comes back fast. Genuinely learned material re-forms quickly with a prompt. Material that was only ever memorised has to be built from scratch again, because there was never any structure to rebuild from.

So the real question isn't "can they recall it?" It's "when we go back to it, does something click, or are we starting over?"

And be careful with your reaction. If a child learns that admitting confusion produces anxiety in you, they'll stop admitting it — and you'll lose your only reliable instrument.

What should I do if my child is coasting?

Resist the urge to add volume. More of the same practice mostly makes the pattern-matching stronger.

Add difficulty of the right kind. Not harder numbers — less scaffolding. Fewer worked examples, more "have a go before I show you." Struggling productively before being told the answer is uncomfortable and unusually effective.

Interleave. Mix topics rather than doing thirty of the same question. Thirty identical questions teach the shape. Mixed questions force the child to identify which idea applies, which is the skill that actually transfers.

Space it out. Revisit old material deliberately, at intervals. This feels inefficient and is one of the best-established findings in learning research.

Value the explanation over the answer. If the only thing that gets praised is a correct answer, that's the only thing that gets optimised for. Ask how they got there. Be visibly more interested in that than in the tick.

How does Skoolio measure understanding differently?

Skoolio doesn't produce a grade, because a grade would answer a question we don't think is the useful one.

Progress is mastery-based: a concept isn't marked as learned because a child got it right once, but because they've demonstrated it holds up — across time, across formats, and when they have to explain it rather than reproduce it. Nothing moves on until it's genuinely understood, and nothing is capped because of where a child started.

Several of the learning modes exist precisely to catch the gap this article is about. Teach Me has the child explain the concept back, which pattern-matching cannot survive. Debate requires holding a position and defending it, which needs structure rather than recall.

And the parent dashboard shows you evidence of understanding rather than a score — what clicked, what's shaky, what they gravitate towards.

You asked whether the tests are telling you the truth. That instinct is a good one. Trust it, and go and find out directly — the answer is usually more encouraging than the worry, and where it isn't, you've found something worth fixing while it's still small.