Nationwide UVA-led study finds Montessori preschool boosts learning at a lower cost

When the University of Virginia psychologist Angeline Lillard began studying Montessori education more than 20 years ago, she did not expect to find that approach to produced large, measurable differences in children’s development. But a growing body of evidence — capped by a new nationwide study published in Proceedings of the National Academy of Sciences — has changed her mind.

“I didn’t believe that this was going to be the outcome,” said Lillard who is the College and Graduate School of Arts & Sciences’ Commonwealth Professor of Psychology, “But I’ve become convinced by the data over time.”

The new research, conducted in partnership with the American Institutes for Research (AIR), and also co-authored by grad student Emily Daggett, a third-year student in UVA’s Department of Psychology, and David Loeb, a fifth-year graduate student at the University of Pennsylvania, is the first randomized,   lottery-based study of public Montessori preschool programs across the United States. The study followed 588 children from age three through the end of kindergarten who had applied to 24 public Montessori schools in Colorado, Connecticut, Maryland, New York, Tennessee, Texas, Wisconsin, Virginia and the District of Columbia.

A Rigorous, Nationwide Study
“This was the first study to be nationwide, and it’s really rigorous,” Lillard said. “We did not handpick schools; we went to every public Montessori school in the country that admitted children at three by lottery and asked them to participate.”

Using school lotteries allowed the researchers to compare children who won admission to Montessori programs with those who did not, creating a rare opportunity to examine cause-and-effect relationships in early childhood education. The team analyzed the data multiple ways — including both “intention-to-treat” and compliance-based approaches — while controlling for factors such as socioeconomic status, race, gender and home language as well as scores at the start of the study (PK3).

“No matter how we did it, we got a consistent story,” Lillard said.

By the end of kindergarten, children admitted to Montessori programs showed significantly stronger outcomes in reading, short-term memory, social understanding, and executive function — which includes skills like planning, attention control, problem solving and self-regulation. Math outcomes also showed positive trends in most analytic models.

What stood out to David Loeb, the study’s lead quantitative analyst, was the strength of the findings.

“Because of the rigor of the study, they’re definitely the strongest findings to date on the effectiveness of Montessori education,” said Loeb, who conducted the analysis as an Institute of Education Sciences predoctoral fellow while completing his Ph.D. at the University of Pennsylvania.

Loeb said the scale of the study posed substantial challenges, from managing data collected across four waves to addressing missing information — an issue common in large, longitudinal studies of young children. The research team used advanced statistical techniques, including multiple imputation and extensive sensitivity analyses, to test whether alternative explanations could account for the results.

“Really none of them seemed to have any sort of meaningful impact,” Loeb said. “There just wasn’t much evidence that these other factors were likely to explain what we found.”

Stronger Outcomes at Lower Cost
In addition to academic and developmental outcomes, the researchers at AIR conducted a comprehensive cost analysis comparing Montessori and conventional public preschool programs.

“Over three years, every child in Montessori versus in conventional preschool costs their districts, on average, $13,127 less,” Lillard said.

The savings are driven largely by Montessori’s mixed-age classrooms and greater reliance on peer learning, particularly for 3- and 4-year-olds. Lillard emphasized that lower cost does not mean lower quality.

“Montessori intentionally has higher student:teacher ratios at ages three and four, and the materials support the learning,” she said.

Looking Ahead
With the largest Montessori lottery study to date now complete, Lillard and her collaborators are focused on what comes next.

“One thing we really want to know is what happens as they go on,” she said, noting plans to examine outcomes as children move into later elementary grades.

Loeb agrees the findings raise important new questions.

“We found that the effects didn’t really emerge until the end of kindergarten, which is a very different pattern than most preschool studies,” he said. “Understanding why that happens could tell us a lot about how the Montessori model influences development.”

For Lillard, the accumulation of evidence has shifted the broader conversation about early childhood education.

“You’ve got to start paying attention at one point,” she said. “The data — not just mine, but what other people are finding as well — are becoming really convincing.”