Why Does My Child Get Math Right in Class and Fail It on the Test?

Last updated: August 19, 2026

Three cognitive scientists set out to settle an argument that had run through classrooms for decades. Is it better to let children discover ideas on their own, or to teach them directly, step by step? Your child is not broken, and they are not lazy, whatever a hard math test seemed to say last week. But the 2006 review that weighed in, by Kirschner, Sweller and Clark, was blunt. For a learner who is new to a subject, minimal guidance does not work. It teaches less, it teaches slower, and it quietly convinces a struggling child that math is the one place they will always come up short.

TL;DR

Kirschner, Sweller and Clark (2006) found that for a learner new to a subject, minimal-guidance discovery teaches less, and slower, than explicit, guided instruction.
Left to discover math alone, a struggling child often decides they are "no good at math" and picks up misconceptions that are harder to unlearn than to teach correctly.
The "knew it yesterday, failed the test" pattern is prompt dependence, when a child did the math with help and never practiced it alone.
Good instruction follows three steps. The adult models it, the pair practice together, then the child does it independently, and that last step is the one usually skipped.
The discovery approach survives as a systems failure, few math specialists and a research-practice gap, the same shape as reading's guessing-from-pictures era.

FROM THE VIDEO

Key moments from Debunking Math Myths: Truths Every Educator Should Know (Ep 28) with Brendan Lee and Heather Brand:

Why letting a struggling child discover math on their own teaches them they are “no good at maths.” Watch at 06:44
The reason a child does every problem right with help and then fails the test. Watch at 22:23
The three-step sequence good instruction follows: model it, practice it together, then do it alone. Watch at 13:19

Common questions from parents

My child understands math in class but fails the test. Why?

This is usually prompt dependence, not a memory problem. Your child had learned to do the math with small helps in place, a hint, a worked example on the page, an adult nearby, and the test was the first time they tried it with none of those. Doing it in the moment is performance. Learning is what is still there when the help is gone. The fix is more practice doing it alone, not different or easier work.

Isn’t letting children discover math themselves better than direct teaching?

For a child who already has a strong foundation, wrestling with an open problem has real value. For a child who is new to the material, the evidence points the other way. Kirschner, Sweller and Clark (2006) found that minimal-guidance instruction teaches novices less, and slower, than explicit, guided teaching. Discovery is not wrong everywhere. It is the wrong starting point for a child who is still missing the basics.

My older child still counts on their fingers or uses blocks. Is that a bad sign?

No. What a child needs is set by what they have already built, not by their age. Math is highly hierarchical, so a child who missed an early foundation stays stuck at that spot regardless of grade. Keeping the blocks while the understanding is still forming is sensible, as long as the hands-on work is tied out loud to the numbers it stands for.

How do I know if this is dyscalculia?

Dyscalculia is a specific math learning difference, estimated to affect somewhere between 3 and 7 percent of children depending on how it is defined. The prompt-dependence and foundation gaps described here affect many more children than that, and they respond to teaching. A screener is a starting point, not a diagnosis. If your child might need formal accommodations at school, such as an IEP or 504 plan, or you suspect a vision, hearing, or medical cause, pursue a professional evaluation too, since that is the only route to those supports.

Your child got every problem right last night and failed the test today. It is not laziness. It is prompt dependence: they did the math with help and never once practiced it alone. Tweet Now Follow Learning Success

The “Let Them Figure It Out” Idea Has It Backwards

The idea sounds wonderful. Hand children a rich, open problem, put them in groups, step back, and let them discover the math for themselves. It feels warmer and more modern than an adult at the front explaining how something works. For a child who already has a strong foundation, there is real value in wrestling with a problem alone. The trouble starts with the child who does not.

This is the heart of what Kirschner, Sweller and Clark found. A novice, someone new to the material, does not yet have the knowledge to steer their own discovery, so they flounder. That paper has been cited thousands of times, and the direction of its finding is not in serious doubt. For learners who are new to a subject, explicit, guided instruction teaches more, and faster, than leaving them to work it out. No one is arguing that rich problem-solving has no place. The point is narrower. It is the wrong starting place for a child who is still missing the foundations.

Here is what that costs a struggling child, in the words of former teacher Brendan Lee. Left to discover math unguided, they mostly work out one thing. They are, they decide, “no good at maths”. The confident kids run ahead. The child who is behind spends the lesson quietly deciding the subject is not for them. Worse, the misconceptions they pick up are harder to unlearn than they would have been to teach right the first time. “I am bad at math” is not a description of where your child stands. It is a prediction they are making about where they are going, and every lesson either confirms it or rewrites it.

Author Quote "

Doing something right in the moment is performance. Learning is what is still there a week later with no help.

Laura Lurns Learning Success Expert

"For a learner new to a subject, minimal-guidance instruction produces less learning than explicit, guided teaching." (Kirschner, Sweller & Clark, Educational Psychologist, 2006)

Why “They Knew It Yesterday” Ends in a Failed Test

You sat with your child last night and they got every problem right. Today the test comes back failed, and the teacher is baffled because “they showed me they could do it.” There is a name for what happened. Your child had become prompt dependent. They could do the math with a nudge from you, a hint, or a worked example on the page. The test was the first time anyone asked them to do it with none of that.

Doing something right in the moment is performance. Learning is what is still there a week later with no help. Good instruction closes that gap on purpose, in three moves. The adult models it and thinks out loud. Then the pair practice it together, with constant checking. Then the child does it alone. That last step, independent practice, is the one that gets skipped when a lesson runs out of time, and it is the one that does the learning.

This is also why homework that goes fine with you sitting there tells you so little. A few hidden prompts fake independence completely. “Remember to carry the one” as you walk away is a prompt. A solved example left at the top of the worksheet is a prompt. Two children doing “independent” work as a pair is still guided practice, because there is no way to tell who is leaning on whom. If you want to know whether your child has it, step back, stay quiet, and watch them do one from a blank start.

Key Takeaways:

Discovery backfires for novices: Minimal-guidance learning teaches a struggling child less, and slower (Kirschner, Sweller & Clark, 2006).

"Knew it yesterday" is prompt dependence: A child who only does math with help fails the first unassisted test.

It is a systems problem: Few math specialists and a research-practice gap keep a debunked approach alive.

Math Has Its Own “Sold a Story”

Reading has already been through this reckoning. For decades, popular programs taught children to guess words from pictures and context, until eye-tracking research showed that skilled readers do the opposite. They process nearly every letter, and guessing turns out to be the habit of the struggling reader. A national investigation named “Sold a Story” put it in front of parents in 2019, and states are still rewriting reading laws because of it. Math has its own comfortable, debunked idea, and it is the belief that children learn it best by discovering it themselves.

None of this is a failure of your child’s teacher. It is a failure of the system behind them. Few primary schools have a math specialist. The teacher shortage puts non-specialists in front of math classes. And teacher training rarely hands anyone the specific knowledge of how to teach math well. So teachers reach for the materials they are given, and the romantic idea that children should discover it all keeps circulating. We have seen this shape before. Researchers tested “learning styles” in 2008 and found no evidence that matching teaching to them helps, and a 2024 analysis reached the same verdict. Yet a 2020 review of educators across 18 countries found nearly 9 in 10 still believe in it. That is not a science problem. It is a systems problem.

The kinder, harder alternative is the one the evidence points to. Teach it directly, build the foundations in order, and give a struggling child the practice they were denied. That includes plain fact fluency. When a child does not have to burn working memory computing three times four, that memory is free for the interesting problem in front of them. It includes letting an older child keep using blocks or their fingers when that is where their understanding sits. What a child needs is set by what they have already built, not by their age. And it means treating the number sense under the math as a skill built one honest step at a time.

Author Quote "

“I am bad at math” is not a description of where your child stands. It is a prediction they are making about where they are going.

You want your child to walk into math believing that effort pays off, because it does, and to stop hearing a failed test as a verdict on who they are. What stands in the way is not your child, and it is not their teacher. It is a comfortable, long-debunked idea, that a struggling learner will find their own way to the math if only we step back, kept alive by a system that never equipped teachers to do the harder, better thing. The people who spend the most hours watching your child struggle and succeed are not the ones who wrote that idea into the curriculum. They are the ones reading this sentence right now.

The Brain Bloom program builds the underlying skills math draws on, with you as the coach and a clear, guided sequence in place of guesswork.

But a math struggle rarely travels alone. A child working this hard at numbers often also carries a heavy working-memory load, an attention struggle, or the anxiety that grows every time math has felt like failure. All-Access opens with a whole-picture assessment that looks across every processing system, so you begin with the skill that actually needs building first. Start your seven-day All-Access trial and find out where your child’s real gap is.

Why we use AI, plainly: it writes from a knowledge base our team maintains and audits—we work through it line by line and pull anything the evidence stops supporting. Doing that by hand for every family would take weeks each. This way the roadmap you get on Tuesday reflects what we corrected on Monday, and a human still reads it before you do.

Your school district must evaluate your child free of charge if you ask in writing, whatever your income and whatever the outcome (US, 34 CFR 300.111 and 300.301(b)). That route takes time and answers a different question than you do. This one starts today, from what you already know.

You get it FREE with your trial.

Get My FREE Assessment & 7-Day Trial

Your answers stay yours. We do not sell your personal information, and we do not hand identifiable assessment data to outside AI companies to train their models. See our privacy policy.

Cancel anytime during your 7-day free trial—no risk.

A screener is a starting point, not a diagnosis. If your child might need formal accommodations (an IEP or 504 plan), or you suspect a vision, hearing, or medical cause, pursue a professional evaluation too—that's the only route to those supports.