Montessori: what the research really shows, and what it changes for running a school

What the recent research shows

The study published on October 20, 2025, in PNAS (Lillard, Loeb, Berg et al.) followed 588 children who entered 24 public Montessori schools in the United States by lottery, from age 3 through the end of kindergarten. At ages 3 and 4, Montessori and non-Montessori children had comparable results. The difference appeared at the end of kindergarten: Montessori children had significantly higher scores in reading, short-term memory, theory of mind (understanding other people's intentions) and executive function (planning, controlling attention, self-regulation).

The result is notable because it runs against the usual pattern: most preschool programs show immediate gains that fade by the end of kindergarten. Here, the benefits appear precisely when the others disappear. The study works on an intention-to-treat basis: it compares all children who won a place with those who did not, which limits selection bias.

Two earlier studies, often cited, already pointed in this direction: the 2006 Milwaukee study (Lillard and Else-Quest, in the journal Science) found higher results in reading, mathematics and social cognition at age 5, and a 2017 study (Lillard et al., Frontiers in Psychology) showed even more pronounced benefits for children from low-income families.

What the research does not say

First limitation: cost. The budget analysis in the PNAS study concludes that three years of public Montessori (ages 3 to 6) cost about 13,127 dollars less per child than a traditional program, mainly thanks to higher child-to-adult ratios at ages 3 and 4. A counterintuitive result, but one specific to the American public-school context.

Second limitation: the study covers American public schools with high Montessori fidelity, meaning they genuinely apply the principles. Yet the word Montessori is not protected by law anywhere: any school can use it. Federations (AMI, AMS, and in France the Association Montessori de France) serve precisely as markers of fidelity. A study of Montessori schools of uneven quality would yield different results.

Third limitation: research measures average effects across whole populations. It says nothing about what happens in a given classroom on a given day. For a head of school, these results shed light on the legitimacy of the pedagogy; they do not replace day-to-day observation.

Three principles, three administrative consequences

Mixed ages. A 3-6 classroom brings together three levels, each moving at its own pace. Consequence: tools that think in terms of grades and age classes (end-of-year reports identical for everyone, lists by level) fit badly. What needs tracking is each child's progress across a three-year cycle.

Materials and presentations. Montessori progress can be read in the presentations given, to revisit, mastered. Consequence: the relevant record is not an end-of-term grade but a history per child and per activity, linked to the qualities it develops (concentration, independence, order, coordination, fine motor skills, social skills, self-confidence, perseverance).

Independence. The child chooses their work within a prepared environment; the teacher observes and guides. Consequence: the value of record-keeping comes from its regularity and precision, not from its weight. A tool that takes one minute after a presentation gets used; a six-page termly form does not.

The French landscape: outside the state contract, but not off the radar

In France, the vast majority of Montessori schools are private establishments outside the state contract, known as hors contrat, meaning they receive no public funding. According to the association Créer son école, 85 hors contrat schools opened for the 2026 school year, bringing their total to 2,665, while 44 closed. Hors contrat education remains a minority, about 1% of pupils, but it is growing, and alternative pedagogies are among the new openings.

Outside the contract does not mean without obligations: declaration to the regional education authority, checks on the acquisition of the common core, safety, registers, communication with families. For a team that is often small (a head who also teaches, sometimes a part-time administrator), these obligations pile up on top of the pedagogical work itself.

This is the exact point the research does not measure but that management reveals: a pedagogy that follows each child over three years calls for a tool that does the same. Spreadsheets and software designed for conventional schooling force you to work around their categories: groups by age, numerical grades, watertight school years.

What a suitable tool must do, concretely

Nothing extraordinary, but the right things: a program structured by domains and subsets, in place from the moment the workspace is created (not a month of configuration); tracking per child and per presentation, linked to the expected qualities; mixed-age groups as the default mode, not as an exception; and ordinary management (scheduling, attendance, families, staff hours) in the same place, so that the head does not lead two separate lives: the life of pedagogy and the life of spreadsheets.

That is the bet Popinz makes with its Montessori pack: the four domains of the 3-6 program, twenty subsets, 174 presentations and eight expected qualities are set up automatically in a new workspace. The Montessori demo is open online: you can judge on the evidence rather than on promises.