I Tried Blending Montessori, Challenging Problems, and Learning Science—Here’s What I’d Change
On Tuesday morning, Nora, a seven-year-old, sat on the rug with a set of golden beads spread in front of her. She’d been building numbers for about twenty minutes, carefully lining up the thousand cubes and hundred squares. I watched her pause, then reach for a small whiteboard where she’d written “4,832 + 1,169.” This was not a Montessori material. I’d added it myself—a challenging problem, the kind you might find in a math circle or a puzzle book. She attempted to add by carrying over, using the beads to represent the numbers. But she mixed up the thousand cube with the hundred square and erased everything in a huff.
I wanted to jump in. The learning science part of my brain said, “She needs feedback now.” The Montessori part said, “Let her work through the error.” The challenging problem part said, “That was too hard for her.” I stood there, not helping, and that moment became the reason I started rethinking the entire framework.
A few weeks earlier, I had decided to combine three things I loved: Montessori’s self-directed, hands-on approach; the kind of hard, non-routine problems that kids rarely see in school; and the practical findings from cognitive science about how learning actually sticks—things like retrieval practice, spacing, and interleaving. I made a simple draft framework. Kids would choose their own work from a prepared environment, but three times a week I’d introduce a challenging problem that required them to use what they’d been exploring. I’d also add periodic “recall” sessions where they’d write or draw what they remembered from past weeks.
It sounded solid. It turned out to be messy in all the ways that weren’t obvious from reading.
Nora’s addition problem was a prime example. She had chosen to work with the beads on her own, which felt like a Montessori win. But my “challenging problem” was not aligned with her current understanding. She knew how to do addition because we’d practiced with stamp games and bead frames, but she didn’t have the concept of regrouping across three places yet. I had given her a problem that required that step, assuming that struggle would teach it. Instead, it just made her feel like she was bad at something she had actually been enjoying.
That evening, I looked at my notes. I had written, “Use problems that are just outside reach.” That’s true in theory. But I hadn’t thought about what “just outside” meant for a specific child. Nora needed a problem that used only two-digit numbers, or one that gave her the option to work with a partner. She also needed me to stop hovering. When I sat down next to her and said, “You don’t have to finish that now,” she looked relieved. She put the beads back and chose a map puzzle instead.
I realized my framework had missed something important: the child’s own sense of timing. Montessori allows the child to choose when they’re ready. Learning science says that struggling with a problem is necessary for deep understanding. But both of those only work when the problem is meaningful to the child. A challenging problem can’t just be hard. It has to be a problem the child cares about solving.
So I adjusted. Instead of a fixed “challenge of the day,” I started keeping a shelf of problems in card boxes, each one color-coded by difficulty and attached to the material the child was already working with. A child using the bead frame might find a card that asks, “Could you use the bead frame to find out how many marbles you’d need to share equally among three friends?” Another card might say, “Tell a friend what you did yesterday and ask them to write the numbers.” These were not deep math problems, but they were designed to make the child connect what they were doing with a question that had no immediate answer.
That helped some. But a new issue emerged.
I had a third grader named Marcus who loved reading facts about space. He’d spend his whole work period at the library corner, flipping through an encyclopedia of planets. Under my framework, I wanted him to engage with retrieval—to remember and apply what he was learning. So I started giving him a “recall sheet” on Fridays. He’d have to write five things he remembered from the week’s reading. The first week he wrote: “Neptune has 16 moons. Mars has 2. Jupiter has 95.” Then he stared at the paper and said, “Is that enough?”
That was when I saw the second flaw. Retrieval practice works best when you’re pulling from memory to build understanding, not just listing facts. Marcus was just listing facts. I hadn’t given him a reason to use the information. So I changed the recall sheet into a problem: “Design a new planet that you could live on. Use at least three facts you learned about a real planet to explain what it would be like.” That made him think. He wrote about a planet with 95 moons where you could always see at least one in the sky. He remembered the fact because he’d made something with it.
But that also made the Montessori part more complicated. Marcus had chosen his reading on his own, but now I was imposing a task on it. He didn’t seem to mind—he liked the planet design—but I could see that for other kids, my “application” requirement would feel like a test. I had to be careful. I started offering the design challenge as a choice, not a requirement. Some kids did it. Some just wanted to read. That was okay.
The third thing I changed was the hardest. I had assumed that mixing materials from different subjects would naturally create interleaving—the learning science idea that switching between problem types improves retention. But in practice, kids would often stay with one material for weeks. That’s partly what Montessori encourages: deep focus, repetition, mastery. Interleaving asks you to shift before you’re too comfortable. Those two ideas clashed more than I expected.
I tried to solve it by creating a “menu” of review activities that kids could rotate through during the last twenty minutes of the afternoon. One day was bead work, one day was math problems, one day was writing about a science idea. It wasn’t elegant. Some kids balked. But it did produce a subtle shift. When we brought back old topics, kids initially fumbled, then remembered more than they expected. That fumbling felt productive, even if it didn’t look polished.
At this point, my draft framework is still in a spiral notebook with a lot of crossed-out lines. What I keep is the basic instinct: let kids choose, push them with problems that make them think, and teach them how to remember. What I’ve added is flexibility. I’ve learned that a challenging problem is not a fixed thing—it depends on the child, the material, and the moment. I’ve learned that retrieval is only meaningful if it’s connected to something the child wants to think about. And I’ve learned that Montessori and learning science aren’t enemies, but they do ask different questions. Montessori asks, “What is this child ready for?” Learning science asks, “How do we make this stick?” The answer, at least for me, is to listen to the child first.
Tomorrow, Nora will probably go back to the golden beads. I plan to set out a card with a smaller addition problem—still two-digit, but only one regrouping. I’ll put it next to the bead frame, not on her mat. If she sees it, she’ll probably try it. If she doesn’t, I’ll leave it. I’ll let her show me what she’s ready for. That feels closer to right.
Please indicate: Thinking In Educating » I Tried Blending Montessori, Challenging Problems, and Learning Science—Here’s What I’d Change