At a glance
Key points
- Across 24 studies, combined academic results favored Montessori by 0.24 standard deviations, a small average difference.
- Across 18 studies, combined non-academic results favored Montessori by 0.33 standard deviations, a small to moderate average difference.
- The review found no clear difference for science, social studies, creativity, or social skills because the uncertainty ranges included no difference.
- The studies covered preschool through high school in public and private settings, but coverage was thin for some ages and outcomes.
- The authors searched for studies through February 2020, so the review does not include later research.
01Results by subject and outcome
The Campbell Collaboration review combined results from 32 studies whose Montessori and comparison groups had similar measured starting points. Across 24 studies, the combined academic difference was 0.24 standard deviations in favor of Montessori. Researchers often call this number Hedges' g. In practical terms, it lets them compare results measured with different tests. The review authors rated the evidence for the combined academic result as moderate and reported substantial variation among studies. Across 18 studies, the combined non-academic difference was 0.33 standard deviations in favor of Montessori, also with substantial variation. These are averages across studies, not the result a family can expect from one school.
Language and literacy favored Montessori by 0.17 standard deviations across 16 studies, while mathematics favored Montessori by 0.22 across 12 studies. General academic ability favored Montessori by 0.26 across nine studies. The authors described those average differences as small in ordinary statistical terms. Science drew on three studies and produced an estimate of 0.15, but its uncertainty range ran from a result favoring conventional education to a result favoring Montessori. Social studies drew on one study and had an estimate near zero. Neither result established a clear difference.
Executive function favored Montessori by 0.36 standard deviations across 11 studies. Students' reported experience of school favored Montessori by 0.41, though the authors urged caution because few independent study groups supported that estimate. Creativity had an estimate of 0.26 and social skills an estimate of 0.23, but both uncertainty ranges included no difference. The review therefore supports a positive combined non-academic average while leaving those two individual areas unresolved. A headline that says every measured area improved would misstate the paper.
What to check
- Read the result for the specific area you care about, not the combined headline alone.
- Treat an uncertainty range that includes zero as no clear difference in this review.
- Keep the size of the average difference separate from the quality of your local school.
Sources for this section: 1
02Ages, settings, and gaps in the review
The included studies covered preschool, elementary, middle, and high school in public and private programs. Coverage was uneven. The review found too much uncertainty to conclude that Montessori's average effect differed by grade level. No included high-school study supplied a non-academic result. Twenty-eight of the 32 studies examined full Montessori programs, while four examined a short Montessori addition inside another program. The authors searched for studies published through February 2020. Their result cannot summarize research published after that cutoff or answer a question the included studies did not measure.
The authors tested whether results differed by preschool, elementary, middle, or high-school level. The estimates varied, but the tests did not establish a reliable grade-level difference. Sparse evidence made some comparisons imprecise. This matters when you apply the review to a local decision. A combined result spanning several ages cannot tell you that a particular program for one age group has the same average. Look for studies that match the age and outcome in your question before narrowing your school list.
The review included public and private schools and reported larger averages for private programs in some analyses. The authors called that evidence preliminary because the estimates were imprecise and program differences may explain part of the pattern. Private status does not identify the cause. Tuition, admission, staff stability, program completeness, and family selection can differ at the same time. Use the overall review to judge the method's research base, then ask a local school to show how its classroom works.
What to check
- Match the age group and school type in a study to the decision you are making.
- Check the study-search date before treating a review as current.
- Treat public-private comparisons as preliminary unless the analysis separates the differences between programs and families.
03Why the study design still matters
Families who choose Montessori can differ from comparison families before school begins. The review included studies that used random assignment or statistical methods to create similar starting groups. That rule removes many weaker comparisons, but it cannot make all 32 studies equally strong. Researchers still had to handle families leaving a study, assigned places that were declined, differences in classroom practice, and measures collected over different periods. The review authors also reported substantial variation among study results. Study design tells you how much confidence to place in an average, while a school visit tells you whether your local option resembles the program studied.
Random assignment and admission lotteries can reduce differences caused by family choice when researchers follow the groups as assigned. Statistical matching uses the information a study collected, so an unmeasured difference can remain. Ask how each group formed, how many participants left, and whether the analysis followed the original groups. You do not need to master every formula. These three questions reveal whether the comparison started on firm ground and stayed intact.
Compare the local school with the features described in the research. Ask about staff preparation, the daily schedule, how lessons connect, the range and condition of materials, and how long families tend to stay. Then add your own requirements for travel, hours, fees, access, and communication. A small average difference cannot compensate for a program your family cannot attend or a local classroom that uses the Montessori name without its central structures.
What to check
- Ask how the researchers formed the groups and who remained in the final result.
- Compare the program in the study with the school you can attend.
- Use fees, hours, travel, staff continuity, and classroom practice in the final decision.
Before you decide
Questions to use
- Which outcome did the study measure, and how large was the reported difference?
- Did the uncertainty range include no difference?
- Which ages and types of school supplied the result?
- How did researchers form the Montessori and comparison groups?
- Does the local school use the main structures described in the study?
Sources3 sources
Montessori education's impact on academic and nonacademic outcomes · Campbell Collaboration
2023 review with outcome-level effect sizes, uncertainty ranges, age and setting analyses, study-search dates, and evidence ratings.
Montessori education: a review of the evidence base · Chloe Marshall
2017 review of Montessori ideas, differences in how programs work, research findings, and limits in the evidence.
A national randomized controlled trial of public Montessori preschool · PubMed record
Describes a national public-preschool trial that assigned places by chance, allowing a stronger comparison between groups.