Skip to main content

When Do Kids Decide They're "Not a Math Person"?









Nov 17, 2026 · @MrPurdyMath

I've heard it at parent conferences, in PD sessions, and from colleagues in the teachers' lounge: "I was never a math person." Sometimes it's said with a laugh. Sometimes it's said as an apology. And far too often, it's said by a parent standing right next to their kid.

Last week, Public News Service ran a piece featuring Jeffrey Houdek, a University of New Mexico lecturer who prepares future teachers to teach math (via KSFR). He made a point that stopped me: families rarely tell their kids "our family isn't good at reading," but they say the math version all the time. We've made it socially acceptable to opt out of math. And kids are listening.

So when does that opting out actually start? A study published this fall gives us a clearer picture, and I think it points squarely at elementary school.

What the research shows

In the September/October 2026 issue of Child Development (published online in April), Sirui Wan, Jing Tian, Siling Guo, and Fani Lauermann looked at four large U.S. longitudinal datasets following students from roughly age 6 to 18 (read the study). Each year, students were sorted into one of four groups based on how important they rated math and language, or how capable they felt in each: a "math person," a "language person," "both-high," or "neither-high."

A few findings jumped out at me:

  • The "both-high" kids disappear, starting in first grade. The share of students who valued (or felt good at) both math and language shrank steadily from Grade 1 through Grade 12, by about 2.9 percentage points a year for importance ratings. Kids begin sorting themselves almost as soon as they start school.

  • The "language person" group grows the whole way through, by roughly 2.2 percentage points per grade on importance ratings, and faster for girls than boys.

  • The "math person" group grows through elementary and early middle school, then levels off and dips slightly in high school. In other words, elementary and early middle school is the window when kids are still joining the math group. By high school, that door is mostly closed.

  • It matters later. In the two datasets with adult follow-up, students who were "math persons" in high school went on to hold more math-related jobs (in one sample, 44.86 vs. 39.19 on the study's math-intensity measure).

A note on what this study can and can't tell us. This is correlational research, so it doesn't prove that any particular teaching practice causes kids to drift away from math. The datasets are U.S.-based and predominantly white, each identity was measured with a single question, and some trends (including the math-person curve for ability self-concept) weren't clearly significant on every measure. I'm not going to overstate it. But the overall shape is consistent: identity sorting starts early, and the years when math identity is still growing are the elementary and early middle grades.

That's the window. So what are we doing with it?

Three things we need to change

1. Stop teaching algorithms before understanding

Too many kids meet math as a set of steps to memorize: carry the one, flip and multiply, cross-multiply. When the procedure comes first, the kids who can memorize quickly get labeled "good at math," and everyone else learns they aren't.

But a child who understands why we regroup in subtraction, who can model it with base-ten blocks, draw it, and explain it, owns that idea. When the algorithm comes later, it's a shortcut for something they already understand, not a magic trick they're hoping to remember. Houdek makes the same point from a different angle: each new concept builds on the ones before it, so a gap in understanding in third grade turns into a wall by sixth.

Conceptual understanding first. Then strategies. Then, and only then, the standard algorithm as one efficient option among several.

2. Do away with timed fact fluency

I'll say it plainly: the timed multiplication test may be the single most efficient tool we've ever built for convincing eight-year-olds they're bad at math.

Fluency matters. I want kids to know their facts. But fluency means flexibility, accuracy, and efficiency, not speed under a ticking clock. When we put a timer on it, we're mostly measuring anxiety, and we're teaching kids that being "good at math" means being fast. The kid who's thinking deeply and carefully loses that race every time, and walks away with a story about themselves that can last decades.

Build fluency through number talks, games, strategy work, and lots of meaningful practice. Retire the race.

3. Better prepare our elementary teachers

This is the one that matters most, and the one we talk about least.

Elementary teachers are asked to teach everything, and many came into the profession carrying their own "not a math person" story. That's not a criticism of them. It's a predictable result of the same system I'm describing. But it means we're asking teachers to build confident math identities in kids while some are still working through their own math anxiety.

Houdek argues that elementary teachers need deep foundational math knowledge and strong formative assessment practices, and I agree. That means:

  • Teacher prep programs that teach the math of elementary math, not just the methods.

  • Ongoing, job-embedded PD that lets teachers do math together, struggle productively, and experience the kind of learning we want for their students.

  • Coaching and time to look at student thinking and respond to it, rather than just covering the next lesson.

When teachers feel confident and curious about math, kids notice. When they're anxious, kids notice that too.

What we praise matters too

One more idea from the Houdek piece is worth carrying into every classroom and every kitchen table. Skip "You're so smart." A correct answer often comes from trying a second approach after the first one didn't work. That's what deserves the spotlight.

Try instead:

  • "I noticed you tried a different strategy when the first one didn't work."

  • "Tell me how you figured that out."

  • "Your drawing really helped show your thinking."

Praise the strategy, the persistence, and the reasoning. Those are things every kid can do, and every kid can get better at.

Final thoughts

No child is born a "not a math person." It's a story they learn, often before they're ten years old, from timed tests, from procedures they never got to understand, and from adults who told them, kindly and without thinking, that math just isn't for everyone.

The research suggests elementary school is where that story starts to take hold. That means it's also where we have the best chance to write a different one.

What's one practice you've changed (or want to change) to protect kids' math identities? I'd love to hear about it.


Sources

  • Wan, S., Tian, J., Guo, S., & Lauermann, F. (2026). Do students become more likely to see themselves as a "math person" or a "language person" across K-12? Evidence from four U.S. longitudinal datasets. Child Development, 97(5), 1868–1879. Link

  • Public News Service via KSFR (Nov. 9, 2026). Overcoming a math bugaboo can boost students' lifelong success. Link


Comments

Popular posts from this blog

The Last Pythagorean Theorem Day of the Century

Oct 24, 2026 · @MrPurdyMath Today is October 24, 2026, written 10/24/26. Those three numbers are the side lengths of a right triangle, which makes today a Pythagorean Theorem Day. Soak it in, because it's the last one we'll get this century. The next one isn't until March 4, 2105. I love dates like this because they turn a calendar into a math prompt. Nobody has to be convinced that the date matters. It's right there at the top of every worksheet, every agenda slide and every phone lock screen. All we have to do is ask students to look at it differently. Check the math The Pythagorean Theorem says that in a right triangle, the squares of the two legs add up to the square of the hypotenuse. A date qualifies when month² + day² = year². 10^2 + 24^2 = 26^2 100 + 576 = 676 Picture a right triangle with a short leg of 10, a long leg of 24 and a slanted side of 26. Today's date is that triangle. This is a great moment to slow down with students. Before anyone reaches for ...

What Does Thanksgiving Dinner Cost?

  Nov 20, 2026 · @MrPurdyMath During the short week before Thanksgiving I always struggle with activities. It’s difficult to do new material in a short week, and especially right before a break, because kids are half out the door. During Thanksgiving, families are making grocery lists, and everyone has an opinion about whether the stuffing goes inside the bird. That's exactly the moment for a Data Talk. This one takes ten minutes. It's percents, unit rates and a fairness question, and students bring their own tables to it: whose family has 10 people, who does the cooking, and what "affordable" means at their house. Same bones as every Data Talk: notice, wonder, then ask who the data is about and who it leaves out. The News Behind It Every November the American Farm Bureau Federation (AFBF) sends volunteer shoppers out to price a classic Thanksgiving dinner for 10. The 2026 survey, the 41st, came out November 18, and for the first time since 2022 the price went up. AFB...

AI-Resistant Authentic Learning Experiences: Creating Meaningful Learning Experiences in the Age of AI

AI-Resistant Authentic Learning Experiences: Creating Meaningful Learning Experiences in the Age of AI What if the best way to make learning AI-resistant isn’t to ban AI—but to make learning worth doing? I’m excited to share an article I co-authored with my fellow AMLE Teacher Leader, Marri Bayor, AI-Resistant Authentic Learning Experiences: Creating Meaningful Learning in the Age of Artificial Intelligence. As educators, we’re navigating a new reality. Students have unprecedented access to artificial intelligence, and it’s easy to focus our energy on restrictions, detection, and preventing misuse.  But what if we asked a different question? How can we design learning experiences that AI can’t replace? In our article, Marri and I explore why we believe the answer starts with authentic learning. When students are solving meaningful problems, collaborating, defending their thinking, creating something that matters, and connecting learning to their own lives, AI becomes a tool—not a s...