When Montessori Meets Learning Science: My Draft Framework and What I’d Change
At 9:30 in the morning, Elena was doing exactly what she wasn’t supposed to be doing. She had chosen the golden bead material from the shelf—the tiny glass beads and wooden squares—and laid them out on her mat in a long, careful row. But my plan was for her to work on a challenge card I’d written the night before. It asked: “If you had ten hundred-squares, how would you trade them?” Elena had not touched the card. She was counting beads one by one, slowly, as if I didn’t exist. I hovered for a moment, then sat down on a nearby chair. I didn’t want to interrupt. I also didn’t want to waste the card.
That tension followed me for months. I had been a Montessori assistant for several years, and I’d also spent a lot of time reading about learning science—retrieval practice, spacing, desirable difficulties, feedback. The idea in my head was simple: a Montessori classroom already gives children materials that allow self-correction. Learning science says that struggle is helpful. So why not combine them? I made a vague framework: keep the Montessori environment, add a shelf of “challenge problems” that pushed children slightly beyond their current level, and work in short review questions for things they had already mastered. I called them “challenge invitations” because “problems” sounded too school-like. It was a draft. I told myself I’d adjust.
What happened was revealing. Leo, a nine-year-old, was working with the fraction insets. He had figured out that 1/2 could be shown in different ways but was stuck on whether 2/4 was actually the same size. I knew the answer. I had a strong urge to say, “Put the fourths under the halves and look.” But the draft framework had a rule: if a child is stuck but engaged, don’t interrupt. Let them experience struggle. So I waited. Leo made a pile of fourths, placed them on top of the halves, and said, “Oh. It fits.” The phrase “Oh” was quiet, not excited. He didn’t need me to explain. He needed enough time to make the comparison himself.
There were moments like that. A child trying to multiply 23 by 12 with the stamp game would make an error, see the leftover ten, and fix it without asking. A group of three children spent twenty minutes arguing about how to measure the classroom rug using paper clips, and one of them finally said, “We need a standard—everyone’s hand is different.” None of that came from my challenge cards. It came from the materials, the time, and the trust. But I still thought the framework was doing something. Then the Friday recall block failed.
I had set aside Friday mornings for something I called “remember and show.” I would ask questions like, “Last week, how did the checkerboard work?” or “What did we learn about the water cycle?” It was supposed to be retrieval practice. But Elena looked at me with a flat face and said, “Why do we have to answer questions about work we already did?” Another child started shaking his head and said, “That’s a test.” The word test was not a good word in our classroom. The whole thing fell apart in about four minutes. I wasn’t using Montessori or learning science anymore; I was using a schedule that felt arbitrary. The children sensed it.
What I learned is that retrieval practice can happen without a formal block. I started changing how I gave feedback and what I noticed. If a child finished a piece of work, I didn’t say, “Great job.” I asked a simple follow-up question that required them to retrieve something: “Can you show me one other way to make that amount?” or “What would happen if we started with the quarters instead of the pennies?” The question wasn’t separate from the work. It was part of the work. And if a child didn’t feel like answering, I let them move. It felt less efficient, but it kept the trust intact.
The other thing I changed was my definition of “challenging problem.” In the beginning, I wrote problems that required a clear correct answer: “How many beads are in 10 tens?” “What is the perimeter of the rectangle?” Those worked for some children but often felt like worksheet questions wearing a Montessori costume. The problems that actually engaged children came from the environment itself. A child would ask, “If I have three rulers, why is the carpet still longer?” or “How can I build a tower that doesn’t fall?” Those were the real problems. I started collecting them. I wrote them down in a notebook—not to grade, but to reuse. For example, “How could we measure the carpet using only paper clips?” became a group challenge on a Tuesday afternoon.
Learning science helped me understand why some things worked. Desirable difficulties means that struggle is useful in the right amount. But it also means the difficulty has to be desirable. A child who has no idea what a triangle is will not benefit from being asked to prove that the angles add up. They will just feel lost. On the other hand, a child who has understood fractions will benefit from being asked to compare the pieces in a way that doesn’t look exactly like the last lesson. The challenge is genuinely “just above” what a child can do alone. That’s a very narrow zone, and it changes day to day. Sometimes a child has had a rough morning. Sometimes they haven’t had enough sleep. Sometimes the classroom is too loud. The same child who could solve a complex puzzle on Monday might not be able to start the same puzzle on Thursday.
So I moved from a framework to more of a habit. I observe first. I ask a question only if the child seems stuck and genuinely interested. I do not ask the question if they are moving toward mastery through repetition. Repetition matters—Montessori understood that. Learning science does too, because spacing and practice build memory. But repeating the same work over and over isn’t always wasted time. For a five-year-old, counting beads might be the full challenge, regardless of what I think they’re ready for.
If I could go back and change the earliest draft, I would change a few specific things. I would not set aside a fixed time for challenges. I would instead wait for children to encounter material in their own order, and then add a “what if” after they’ve had time to build confidence. For example, after a child successfully builds a square with the pink tower, ask, “Can you build it starting with the smallest piece on top?” That’s a challenge, but it’s tied to the moment. It also leaves room for the classic “Maybe no, I’m done,” which should be a fine answer.
I would also involve the children more in choosing challenges. Instead of printing cards, I might ask, “What do you want to figure out next?” Some would say “nothing,” and that’s okay. Others would give wonderful questions that are impossible to predict. Those questions are the truest form of challenging problems because they come from the child’s own sense of curiosity. They aren’t “problems” at all, in the worksheet sense. They’re things that feel worth solving.
One last thing I’d change is the word “framework.” It sounds like something formal, with boxes and arrows. For me, it’s become a set of reminders: don’t interrupt; observe enough to know the difference between stuck and lost; ask a question that helps the child, not one that checks them; let unhurried repetition happen; and be willing to throw away any plan that makes a child feel tested.
I still don’t know exactly what to do when Elena spends an entire morning counting beads. Sometimes I think it’s fine. Other times I wonder if I should step in and show her a shortcut. The learning science says retrieval and reasoning are better learned than taught. The Montessori voice says the child has a right to figure it out in her own time. The part of me that worries says, “What if she never reaches a shortcut?” The honest answer is that I don’t know yet. The draft is still a draft. I suspect it always will be.
Please indicate: Thinking In Educating » When Montessori Meets Learning Science: My Draft Framework and What I’d Change