Published 2 June 2026
There is a story we tell children without realising we are telling it. We tell it when we say "I was never good at math" in front of them. We tell it when we say that "some people are just made for math and others aren't." We even tell it when we praise: "you're so smart, you figured that out so fast."
The story is this: mathematical ability is something you are born with, not something you build.
Decades of research suggest this story is incomplete, and the cost children pay for it is higher than most of us assume.
What the research says
Carol Dweck, psychologist at Stanford University, spent decades studying how beliefs about intelligence shape performance. Her central observation: children who believe intelligence is fixed ("I either have it or I don't") tend to give up sooner when they hit difficulty. Children who believe their abilities can grow through effort, strategy, and practice persist longer, look for new approaches, and over time tend to achieve better results.
Dweck describes these two orientations as "fixed mindset" and "growth mindset." Her work has been replicated across many contexts, though academic debate continues about the size of the effects and the conditions under which interventions produce real benefits. The best-supported conclusion is more measured than the popular version: not that talent doesn't exist or doesn't matter, but that talent is not destiny. Sustained effort, the right methods, and a safe learning environment can, over time, significantly reduce initial differences.
At the neurobiological level, repeated practice produces concrete changes: frequently used neural pathways strengthen, and operations that once required conscious effort gradually become automatic. This matters especially for arithmetic: fluency in basic operations (addition, subtraction, multiplication, division) is not an end in itself, but a foundation. When simple operations are automatic, working memory is freed for more complex mathematical thinking.
Sian Beilock, a researcher at the University of Chicago, identified a dimension that is often overlooked: in people with math anxiety, neuroimaging studies (fMRI) have shown that simply anticipating a math task can activate brain regions associated with threat processing and, in some studies, pain. This is not an exaggeration. It is a measurable response with a concrete effect: it can suppress working memory, precisely the cognitive resource a child needs in order to calculate. Sometimes the child does not fail because they cannot think, but because the evaluation environment triggers an alert response that interferes with thinking.
How traditional evaluation can reinforce these beliefs
Traditional math assessment works through timed tests, single grades, and immediate public feedback (right or wrong, in front of classmates). This is not an indictment of teachers or systems: these mechanisms have their own pedagogical rationale. But the evidence suggests that for children prone to math anxiety, they can produce an unintended effect: associating math with exposure and judgment rather than with effort and growth.
A child who makes a mistake at the board, in front of their class, does not always learn from that mistake. Sometimes what they take away is that math is a place where you can be seen to fail.
Jo Boaler, professor of mathematics education at Stanford, documents in Mathematical Mindsets how children as young as six or seven begin to sort themselves: "I'm a math person" or "I'm not." By age nine, this belief is often stable, not because they have uncovered a truth about their potential, but because they have received, repeatedly and indirectly, the message that belonging to one of those categories is fixed.
The difference between practice and testing
The difference is not just semantic. It is structural.
A test has a beginning and an end. It has a score. It has a verdict: you know it or you don't. Every mistake is evidence of not knowing, and possibly a signal that you never will.
Practice carries no verdict. A mistake is the next piece of information: the signal that tells the system what needs to be revisited and strengthened. Practice works precisely through repeated exposure to what is difficult, without the pressure of an immediate result.
Spaced repetition algorithms (SM-2, the Leitner system) are built on this logic: problems you get wrong today come back tomorrow, not as punishment, but as opportunity. The gaps between appearances lengthen gradually as responses become stable. That is how long-term memory consolidates: not through sudden accumulation, but through calibrated return.
When anxiety is not in the way, this is how the brain naturally internalises repeating structures. Basic math is no exception.
What you can do as a parent
A few consistent adjustments can shift a child's relationship with math more than extra tutoring hours.
Watch what you say about your own history with math. "I was never good at this either" or "I could never do numbers" quietly grants permission and models giving up. You don't have to pretend you love math. But you can say: "It didn't come easily to me either. I had to practise a lot."
Praise the process, not the label. Dweck's research suggests that feedback focused on effort and strategy (for instance, "you tried a different approach when the first one didn't work") supports persistence better than "you're so smart" or "you have a gift for this." Well-meaning talent praise can subtly communicate that results should come effortlessly, and that needing to work hard is a sign of weakness rather than the very mechanism of learning.
Let them practise without pressure. A few minutes a day, no score, no comparison. The goal is not immediate performance; it is reconnecting with math in a context that feels safe.
Treat mistakes as information, not failure. "This one hasn't clicked yet. We'll come back to it" is not consolation. It is an accurate description of how memory consolidation works.
What Goldy does differently
Goldy does not turn evaluation into a verdict.
There is an initial placement (necessary to understand where it makes sense to start), but the user does not see a score and does not receive a label. The message is not "your level is X." The message is: "we know where to begin."
When a user gets a calculation wrong, the app does not display a red X, deduct points, or play a failure sound. It registers that the problem needs more practice and brings it back in future sessions, at the interval calculated for optimal consolidation.
Sessions are short (10 to 15 minutes) by design. The evidence suggests that frequency matters more than duration: short, regular sessions tend to produce better retention than long, infrequent ones. For a child with math anxiety, short sessions also mean less cumulative exposure to discomfort.
Goldy does not promise to change everything. It offers an environment where practice is possible without the fear of judgment; from there, progress follows consistency.
Conclusion
Math is not a trait you either have or don't have. It is a skill built the way any skill involving memory, automaticity, and structured thinking is built. You do not need to be a "math person." You need to practise, regularly, without the pressure of a permanent verdict.
If you or your child have come to believe that math is simply not for you, it is not because you have found a true limit. It is because, somewhere along the way, someone (with or without meaning to) taught you that talent is all that counts.
It isn't. And perhaps the most useful thing you can do today is change the story you are telling.
Goldy is a math practice app for children and adults. 10–15 minute sessions, no pressure, no verdict. Start practising for free.