Project Activities
Researchers evaluated the impacts of participating in a Montessori program on students' academic and social-behavioral skills. Through existing lotteries, districts randomly selected students to attend Montessori programs; those who were not admitted were assigned to the waitlist control group for the study. The researchers collected data and conducted analyses to compare outcomes for students who were offered enrollment in a Montessori program with the outcomes for students who participated in the lotteries but were not offered Montessori enrollment (and attended other types of preschool programs instead). They also estimated the cost of implementing the public Montessori preschool model compared with traditional preschool programs in public schools.
Structured Abstract
Setting
The study took place in Montessori and non-Montessori public preschool programs in public, magnet, and charter schools, in Virginia, the District of Columbia, Maryland, Connecticut, New York, Wisconsin, Tennessee, Texas, and Colorado.
Sample
The study sample included 21 public Montessori schools and 662 students from racially and ethnically diverse backgrounds.
This study evaluated the Montessori model for preschool, examining impacts through the end of the 4-year-old year. The Montessori model is based on the theory that young students learn more by action than by thought. Primary classrooms are multiage, including 3- to 6-year-old children. Montessori teachers provide a prepared classroom environment filled with appropriate materials, but students have considerable freedom to select the activities they engage in each day.
Research design and methods
The researchers used a blocked individual random assignment design, where children were randomly assigned within lotteries to either attend Montessori at age 3 or assigned to a waitlist control group within each site. Some participating schools had their own lotteries, and some were part of district-wide lotteries. The researchers followed the same cohort of children for 2 years (fall 2021 to spring of 2023). (The study originally began with a cohort of students entering in the fall of 2019, but the researchers had to cease following those students when their educational experience was interrupted by the COVID-19 pandemic.) The team administered direct child assessments, collected teacher report data, conducted classroom observations, and interviewed district officials about program costs. They also examined the relationship between fidelity of implementation of the Montessori model and impact, estimated the cost of Montessori compared with other preschool models, and shared findings broadly.
Control condition
Students from the Montessori waitlist who are not admitted to the program served as the control group. These students will attend non-Montessori preschool programs or participate in other home care arrangements. The research team tracked students in the control condition and document their enrollment in non-Montessori preschool programs.
Key measures
Primary measures included standardized direct child assessments of children's academic and social-behavioral skills, classroom observations, and measures of fidelity of implementation fidelity. Direct assessments included subtests of the Woodcock-Johnson tests of achievement, the Rubin Social Problem Solving task, a theory of mind scale, a puzzle task measuring mastery orientation, and two measures of executive functioning (the Head Toes Knees Shoulders task and the forwards/backwards digit task). Classroom observation measures included the Classroom Assessment Scoring System and the Developmental Environmental Rating Scale.
Data analytic strategy
Researchers conducted multilevel instrumental variable analyses to estimate the impact of being offered admission to a Montessori school through a lottery, using both Intent to Treat (ITT) and Complier Average Causal Effect (CACE) frameworks. They assessed the impact of Montessori at the end of each of the two years of the study. The study will continue with additional private funding to follow children through kindergarten, at which point the team will use growth curve modeling to examine changes in trajectories for Montessori and control group students.
Key outcomes
- Scores on the Classroom Assessment Scoring System (CLASS), a measure of quality of teacher-child interactions, were statistically similar between Montessori and non-Montessori classrooms observed. Scores in both Montessori and non-Montessori classrooms in the study were generally higher than the national average, suggesting that students in both the treatment and control groups experienced relatively high-quality classrooms (Lillard, A. S., et. al, 2025).
- The researchers found differences in the Montessori preschool environments compared with classrooms control students attended, including in student-teacher ratios, structure, and curricula. Montessori classrooms with larger class sizes (up to 26 students) had higher scores on the emotional support and classroom organization domains within the CLASS than those with smaller class sizes, and classes with higher child-to-adult ratios (up to 13:1) trended towards higher scores on the instructional support domain than ones with lower ratios, making Montessori preschool a less expensive option that obtains equal quality as measured by the CLASS (Lillard, A. S., et. al, 2025).
People and institutions involved
IES program contact(s)
Products and publications
Study registration:
Publications:
ERIC Citations: Find available citations in ERIC for this award here.
Select Publications:
Lillard, A. S. (2019). Shunned and admired: Montessori, self-determination, and a case for radical school reform. Educational Psychology Review, 31(4), 939-965.
Lillard, A. S. (2021). Montessori as an alternative early childhood education. Early Child Development and Care, 191(7–8), 1196–1206. DOI.10.1080/03004430.2020.1832998
Lillard, A. S. (2022). Pretending at hand: How children perceive and process puppets. Cognitive Development, 63, 101202.
Lillard, A. S. (2023). Grand challenges in developmental psychology. Frontiers in Developmental Psychology, 1, 1069925.
Lillard, A. S. (2023). Why the time is ripe for an education revolution. Frontiers in Developmental Psychology, 1, 1177576.
Lillard, A. S., LeBoeuf, L., Borgman, C., Martynova, E., Faria, A. M., & Manship, K. (2025). When bigger looks better: CLASS results in public Montessori preschool classrooms. Early Childhood Research Quarterly, 70, 199-210.
Lillard, A. S., Tong, X., & Bray, P. M. (2023). Seeking racial and ethnic parity in preschool outcomes: An exploratory study of public Montessori schools vs. business-as-usual schools. Journal of Montessori Research, 9(1).
Lillard, A. S.,Tong, X., & Lillard, A. S. (2022). Standardized test proficiency in public Montessori schools. Journal of School Choice, 16(1), 105-135.
Supplemental information
Co-Principal Investigator: Lillard, Angeline
Questions about this project?
To answer additional questions about this project or provide feedback, please contact the program officer.
