How to tell whether your child actually understands something

The fastest way to tell whether your child understands something is to have them explain it to you, out loud, without notes, a few days after the lesson. Nodding, finishing the worksheet and picking the right answer from four options can all happen without understanding. Producing a clear explanation from memory essentially cannot.

Every homeschooling parent has had the sinking experience of a child who sailed through a chapter in October and has no idea what it was about in January. That is not a memory problem. It is almost always a sign that what looked like understanding at the time was something weaker wearing its clothes.

The thing that fools everyone

Psychologists call it the fluency illusion. When material feels easy to process - because you have just read it, because the teacher explained it well, because the worked example is on the same page - your brain reads that ease as "I know this". The feeling is genuine and it is not evidence.

This is why rereading is the most popular study technique and one of the least effective. A large review of study techniques by Dunlosky and colleagues in 2013 rated rereading and highlighting as low utility, while rating practice testing and distributed practice as high utility. Rereading feels productive. That feeling is precisely the problem.

Your child experiences the same illusion, and so do you watching them. Confidence is not a signal.

Check one: explain it back

Ask your child to teach you the idea as though you had never heard it. No notes, no book.

Listen for four things:

• Hesitation at a specific point. Fluent, fluent, fluent, stall. The stall is the gap, and it is usually exactly where the understanding runs out.
• Circularity. "It happens because that is what happens in photosynthesis." Restating the name of the thing as its explanation.
• Borrowed phrasing. Textbook sentences delivered intact. A child who understands rephrases; a child who memorised recites.
• No connections. Each fact arriving separately, with no sense of how they fit.

The research supporting this is strong. Students who explain material to themselves as they study learn substantially more than students who read it twice, a finding established by Chi and colleagues in 1994. Explaining is not just a test of understanding. It builds it.

Check two: ask why, not what

"What are the products of photosynthesis?" tests recall of a list. "Why does the plant need light at all?" tests whether the list means anything.

Good why-questions sound like:

• Why does that happen rather than the alternative?
• Why does this step have to come before that one?
• Why is this true in general and not just in this example?

A memorised answer dies here, because the why was never in the sentence they memorised.

Check three: change one thing

Take the situation and alter a single variable. "What would happen if the plant had light but no water?" "What if the triangle were obtuse instead?" "What if the interest were charged monthly rather than yearly?"

Understanding is the ability to run the model forward in a case you have not seen. Memorisation has nothing to run. This check takes ten seconds and is close to impossible to fake.

Check four: wait, then ask again

Come back in three to seven days and ask about the same idea cold. This is the one most families skip, and the one that separates learning from familiarity.

Roediger and Karpicke showed in 2006 that students who were tested on material remembered substantially more a week later than students who restudied it for the same time, even though the restudying students felt more confident at the time. The confidence and the retention point in opposite directions.

So the delayed check is doing two jobs at once: it tells you what survived, and the act of retrieval makes what survived stick harder.

What a good check looks like

That fourth row is the most valuable thing home education can give a child, and the hardest thing for a classroom to provide: a precisely located gap, fixed on the spot, rather than a grade of 72 per cent with no indication of which 28 per cent is missing.

What each check looks like in four subjects

The four checks are subject-free, which is the point, but what a good answer sounds like differs.

The pattern across all four: a memorised answer states what, and an understood answer states why or how. You can hear the difference without knowing the subject, which is what makes this usable by a parent teaching something they never studied.

Three false signals to stop trusting

"They got 90 per cent." A high mark on work done immediately after the lesson, with the example on the same page, measures short-term pattern matching. The same test in three weeks is the real one.

"They seem confident." Confidence tracks how fluent the material felt, not how well it was learned. The students most likely to be surprised by an exam result are the confident ones.

"They finished quickly." Speed on routine work is good news about fluency and says nothing about understanding. A child who can execute a procedure fast and cannot say why it works has exactly the gap that collapses the following year.

What to do with what you find

A gap found is only useful if the next step is specific.

• Stalled at one step. Reteach that step alone, not the topic. Then have them narrate the whole thing again from the start.
• Right answer, no reasoning. Give them a problem that looks different and works the same way. If they cannot start it, they learned the example rather than the method.
• Fine today, gone in a week. Nothing is wrong with their understanding; the practice was massed. Space it out and make the review a test rather than a reread.
• Fluent but hollow across a whole topic. Go back further than feels necessary. The gap is usually one level below where the symptom shows.

Keep it light

A daily interrogation will make your child defensive, and a defensive child tells you nothing. Some framings that work:

• "Explain it to me, I've forgotten how this works."
• "Teach it to your sister."
• "Pretend I'm your student and I'm being difficult."
• At dinner, with no books out, as a conversation rather than an exam.

The aim is to make explaining something that happens routinely, not something that happens when they are in trouble.

Where TruLearn fits

Explain-back is the core mechanic of what we are building, rather than a feature on the side. TruLearn teaches a concept out loud, then asks the student to say it back in their own words, and grades that explanation criterion by criterion against a rubric written by a person for that specific concept. When a student voices a wrong idea, it is stored in their own words and brought back days later until they get it right.

It teaches AP Biology and introductory college biology today, and there is no parent dashboard yet, so the way you see how it went is still to ask your child to explain what they learned. Which, as this article argues, is the check worth running anyway.

More on the research behind this, or how a session works.