Table 1.

Characteristics of included studies.

Study	Participants	Design	Intervention characteristics	Outcomes/measures	Dates	Funding sources
Alburaidi 2019	62 4th‐grade students in a Middle Eastern (Oman) primary school	Quasi‐experimental PPCG design	Nonspecific Montessori approach; add on to conventional program; use of sensorial learning stations	Science achievement as measured by a researcher‐made test in the curriculum area	2017–2018 academic year	None specified
Ansari 2014	7045 Hispanic and 6700 Black low‐income students in two types of Title‐1 Public preschools in Miami, Florida	Quasi‐experimental PPCG design	Montessori classrooms have serious implementation flaws: single age (4), not the recommended 3‐year age span (3–6). Teachers were Montessori trained	Cognitive, Language, and Fine Motor skills (Learning Accomplishment Profile‐Diagnostic). Parent and Teacher Report of Social Skills and Behavioral Problems (Devereux Early Childhood Assessment)	Not specified	Project funded by the Early Learning Coalition of Miami‐ Dade/Monroe and supported by Grant T32 HD007081‐35 (PI: Kelly Raley) provided by the Eunice Kennedy Shriver National Institute of Child Health and Human Development
Aydoğan 2016	35 preschool children (14 boys, 21 girls) in Turkey	Quasi‐experimental PPCG design	Nonspecific Montessori. The intervention lasted 7 weeks	Children's language development was measured by DVT (Descoeudres Vocabulary Test), VLT (Vocabulary and Language Test), and the PPWT (Peabody Picture‐Word Test—Turkish version of PPVT)	Not specified	None specified
Besançon 2013	N = 80 6–11‐year‐olds. Paris	Quasi‐Experimental PPRM, test‐retest design	Nonspecific Montessori including theatre workshops; serious implementation issue in using just 1 or 2 ages per classroom	Torrance Test of Creative Thinking (Divergent‐exploratory thinking tasks‐toy improvement and parallel lines; Convergent‐integrative thinking tasks—drawing and story)	Not specified	None specified
Coyle 1968	N = 131, 3–5‐year‐olds from 4 schools (low‐income diverse free school, lower‐middle income White neighborhood daycare, upper‐income suburban school with tuition, modified Montessori school with lower‐middle income families who paid a small tuition fee). 72 Montessori students, 59 non‐Montessori	Quasi‐experimental PCG	Private “modified‐Montessori” where 3–5‐year‐olds attend either a morning or afternoon session. Serious implementation issue where classrooms used group instruction. The intervention lasted 12 months with post‐tests occurring at 6 and 12 months; final scores were used here	Five tests of Haptic Perception: Test 1 (identify/name geometric forms), Test 2 (match two geometric forms), Test 3 (match a shape presented in a haptic form to drawing of the same shape), Test 4 (draw a picture of the shape presented haptically), Test 5 (verbally describe the form of the shape presented haptically). Twelve‐month posttest scores used	1965–1967	US Department of Health, Education, and Welfare
Culclasure 2018	n = 7401 third through eighth‐grade public Montessori students, the control group was matched for demographics. In the Montessori group, 55% of students were White, 34% were Black, and 6% were Hispanic. 10% of Montessori students had a special education designation	Quasi‐experimental PPRM	All Montessori schools met a minimum standard of fidelity set by the authors; within those included, half scored as high fidelity in classroom observations, and half scored as low or medium fidelity. Considered a medium implementation overall	Writing, ELA, Math, Social Studies, Science, Executive Function, Social Skills, Creativity	2012–2013 school year to 2015–2016 school year	The Self Family Foundation and the South Carolina Education Oversight Committee
Denervaud 2019	N = 201 children (99 Montessori (42 girls, 57 boys), 102 Traditional (54 girls, 48 boys)) ages 5–13. Switzerland	Quasi‐experimental PCG	13 classrooms in 5 private AMI Montessori schools in Switzerland. Full implementation	Creativity (Convergent, Divergent), Executive Function (Working Memory, Selective Attention, Cognitive Flexibility), Inner Experience of School (Well‐Being), Language/Literacy, Math	Not specified	National Center of Competence in Research, Swiss National Science Foundation
Denervaud 2020	10 traditional and 13 Montessori schools in Switzerland were selected, N = 234 affluent 4–15‐year‐old students (114 girls and 120 boys; 111 Conventional, 123 Montessori). Switzerland	Quasi‐experimental PCG design	13 Swiss Montessori schools meeting AMI structural criteria. Full implementation	Child‐friendly version of the Flanker task: post‐error slowing, post‐error improvement in accuracy	Not specified	The Boninchi Foundation in Geneva, The Department of Radiology of Lausanne University Hospital (CHUV)
Doğru 2015	N = 15 5–6‐year‐olds with ADHD or AD, 6 girls, 9 boys; 8 in the experimental group, 7 in the control group. Turkey	Experimental PPCG	Add on to the traditional curriculum. Montessori “tactile boards, sound boxes, binomial cubes, and color tablets”; each material presented and used for 15‐min sessions/day, 3 days/week, for 2 weeks each (8 weeks total)	FTFK Attention Test (Concentration, test–retest)	Not specified	None specified
Elben 2015	N = 42 6–7 year‐olds in Switzerland (22 girls, 20 boys). 18 control, 17 Montessori including K. 7 new to Montessori not included	Quasi‐Experimental PPCG	Lower elementary bilingual Montessori classes (analyzed children started Montessori by Kindergarten), accredited by the Swiss Montessori Association, private school. Three‐hour work period in AM. Supplementary “specials” (music, P.E., crafts) in PM. Montessori classrooms have 3 teachers and aides for about 30 students. “Several of the teachers have the Zurich teaching diploma and the Montessori elementary teaching diploma, and the others have Montessori diplomas.” The intervention lasted 6 months. Full implementation	ELFE 1–6 (reading comprehension); other reading measures (phoneme—Bako; letter recognition) given but not used (no pretest for latter)	November 2014–May 2015	None specified
Faryadi 2017	180 Malaysian Kindergarteners	Quasi‐Experimental PPCG design	Unspecified Montessori school. Photographs in the article suggest a mix of Montessori and non‐Montessori materials. The intervention lasted 4 months. Medium implementation	Math ability, using a variety of measures to assess components of Critical Thinking, Problem Solving, “responsible learners”	2015	None specified
Fleege 1967	n = 21 Montessori (13 boys, 8 girls) matched for IQ, SES, and a range of other variables with 21 control at a non‐Montessori preschool in the same high‐income Chicago community	Experimental PPCG	Montessori pre‐school; no details given on implementation	PPVT administered at the beginning and end of the school year	1963	Office of Education, US Department of Health, Education, and Welfare
Galindo 2014	600 low‐income bilingual prekindergarten students (300 Montessori, 300 Traditional) ages 4–5 years in Houston, TX. The majority of students in both centers were Hispanic and low‐income (1% White students in Montessori, 0.7% White in Traditional)	Quasi‐Experimental PPCG design	Public bilingual Montessori with 36 classrooms; 94% had some Montessori training, but intervention quality is unknown. The intervention lasted 26 weeks	BBCS:E (Bracken School Readiness) test of concept development administered at the beginning and end of the school year	2012–2013	None specifies
Hoseinpoor 2014	60 preschoolers ages 5–6 in Iran	Quasi‐Experimental PPCG design	Add on to the conventional program. The experimental group went through a series of 12 lessons involving some Montessori materials and practices	Wechsler Intelligence Scale (Attention and Concentration), 4th ed. Wiland Questionnaire of Growth of Social Skills Scale	2013–2014 school year	None specified
Jones 1979	6th and 7th grade follow‐up included 18 who attended M at age 4 (all Black, 75% single parent homes); 24 attended traditional preschool (90% Black; 45% single parent homes); Louisville Head Start study (lower‐SES)	Experimental PPCG	As described in Miller, Dyer, et al. 1975 SRCD monograph. Consultants rated program implementation 6.5/10. Only 4‐year‐olds in the program and teacher training very limited	Vocabulary, Reading Comprehension, Word Study Skills, Math Concepts, Math Computation, Math Applications, Spelling, Language (measures taken from Weschler Intelligence Scale for Children ‐ Revised and Stanford Achievement Test—6th and 7th‐grade points)	1968–1969	None specified
Juanga 2015	Two classes (each n = 120) of Kindergarten students in each of two schools (one private, one public) ages 5–6 years. Not all students at each school took both tests. Philippines	Quasi‐Experimental PPCG	The researcher and assistant implemented Montessori workstations in traditional classrooms with help of a Montessori consultant, focusing on science and math. Length of intervention unclear. It might have been a single day/lesson or add on	Math and science achievement tests created for the study	Not specified	None specified
Kayili 2016a	Montessori (n = 19, 6 girls, 13 boys, avg. 3.68 years), Control Group 2 (n = 16, 11 girls, 5 boys, avg. 3.63 years). Turkey	Experimental PPCG	3–6 Montessori classroom assoc with Selcuk U. in Turkey; other features unclear resulting in unknown implementation classification. A third group was a Montessori classroom supplemented with social skills training composed of 64 lessons; pretest and posttest were around that intervention, with timing unclear	Social (Wally Feelings Test, Wally Social Problem Solving Test)	2013–2014 school year	None specified
Kayili 2016b	63 children 48–72 months old, in Montessori (n = 40, 19 girls, mean age 63.33 mos) or Preschool Education Program created by the Ministry of National Education (n = 23, 13 girls, mean age 63.61 mos) at a single school. Turkey	Quasi‐Experimental PPCG design	See the other Kayili 2016 study; 3–6 Montessori classroom assoc with Selcuk U. in Turkey; other features unclear. Tested at the beginning and end of the school year	Kansas Reflection‐Impulsivity Scale for Preschool	2015–2016 school year	None specified
Kirkham 2017	40 students ages 6–11. 20 Montessori (8 girls avg. age 92 mos, high SES), 20 National Curriculum (12 girls, avg age 97 mos, lower SES). UK	Quasi‐Experimental PPRM design	4 Classrooms in 4 Montessori schools accredited by the UK Montessori Schools Association (MSA) with additional input from the Montessori St. Nicholas Charity. One classroom had fantasy toys (like dolls) available	Test of Creative Thinking—Drawing Production. Groups were equivalent on Expressive Vocabulary Test and Raven's. Also administered a pretense production task	Not specified	None specified
Lillard 2006	N = 112: 5‐year‐olds (25 control (15 boys, 10 girls), 30 Montessori (15 boys, 15 girls)); 12‐year‐olds (28 control (18 boys, 10 girls), 29 Montessori (12 boys, 17 girls)), in Milwaukee WI schools serving a diverse population. The controls were at 27 public inner city schools (40 children) and 12 suburban public private/voucher or charter schools (13 children). Both control and Montessori children had similar family income (average $20,000 to $50,000 annually)	Lottery controlled, PCG	Public Montessori schools recognized by AMI. Full implementation	Woodcock‐Johnson IIIR Applied Problems, Picture Vocabulary, Letter Word, Word Attack, False Belief, Social Problem‐Solving Test, Dimensional Change Card Sort, Playground behavior, School Liking, Narrative task	2005–2006 school year	Jacobs Foundation, Ovid Foundation
Lillard 2017	N = 141 Hartford, CT preschool students, 70 in public Montessori Schools and 71 in Non‐Montessori Schools, followed 3 years. Age 41.15 months avg. at start. 3 cohorts. Half White, half higher income	Lottery controlled PPCG	11 Montessori 3 to 6 classrooms with AMI‐trained teachers in 2 AMI‐affiliated public magnet schools. The intervention lasted for 3 years and the children were tested four times. Full implementation	Woodcock–Johnson IIIR Letter Word, Picture Vocabulary, Math (composite of Applied Problems and Calculation), Theory of Mind Scale, Rubin's Social Problem‐Solving Test, Executive function (Head‐Toes‐Knees‐Shoulders task and Copy Figures from the Visuospatial Processing section of the NEPSY‐II), Mastery orientation (puzzle task), School enjoyment (preference questionnaire), and Creativity (Alternative Uses)	2010‐2015	Brady Education Foundation
Lillard 2012	N = 172 3–6‐year‐olds enrolled in private Montessori and conventional private schools where Montessori parents said would send children if Montessori was not available. Majority White (90%), 5% Black, 3% Asian, 1% Hispanic	Quasi‐experimental PPCG	All teachers in the Classic Montessori group were all AMI‐trained. Tested children at the beginning and end of the school year. Full implementation	Executive function (Head‐Toes‐Knees‐Shoulders, Theory of Mind Scale, Social Problem Solving Task, Woodcock‐Johnson IIIR Applied Problems, Letter Word, Picture Vocabulary	Not specified	None specified
Mallett 2015	N = 1035 students (518 at 2 Montessori schools, 517 non‐Montessori), 1st through 5th grade, approx. 100 at each age level in each group, urban Texas public school district	Quasi‐experimental, PCG	Montessori classrooms had a “full array of specialized Montessori materials.” Montessori teachers also held or were training for Montessori teaching certification. Montessori classrooms had both a teacher and a teaching assistant; adult child ratio same as in all public district schools. 50% of incoming Montessori grade 1 students had no prior Montessori experience. Medium implementation	First and second graders: Iowa Tests of Basic Skills (Reading Vocabulary and Reading Comprehension made up “Reading,” Math Concepts and Math Computation made up “Math”), Third, fourth, and fifth graders: Texas Assessment of Knowledge and Skills. Control for SES, race, gender, and prior year's scores	2011	None specified
Manner 1999	Third‐graders in Broward Country, Florida. Students either went to a Montessori magnet school or a traditional school, both with diverse student bodies. Matched for the third‐grade score in either math (30 pairs) or reading (37 pairs), followed through 5th grade. Within the math group, there were a total of 60 students, 30 pairs, and within the reading group there were 74 students, 37 pairs	Quasi‐experimental PPRM	The study included public magnet Montessori schools with teachers with AMS training completed or in progress. The intervention lasted for three years with the same participants. Medium implementation	Stanford Achievement Tests Total Reading and Math scores. The author only reported combined math across the three years, but reading was separated out	1996–1998	None specified
Miller 1983	Recovered 20 Montessori and 29 Traditional preK Head Start participants in 1978 when in 8th grade; able to view 7th‐grade test scores too. About 90% of the students in the experimental group (students who were assigned a program) were Black and came from low‐income families	Experimental PPCG	See Jones above	Seventh grade: WISC‐R. Eighth Grade: Stanford‐Binet Intelligence Test	1978 follow‐up; intervention in 1965	None specified
Miller 1984	Recovered 20 Montessori and 22 Traditional preK Head Start participants in 1980 when in 10th grade. About 90% of the students in the experimental group (students who were assigned a program) were Black and came from low‐income families	Experimental PPCG	See Jones above	STEP‐Locator achievement tests and the Comprehensive Test of Basic Skills	1979–1980	None specified
Mix 2017	Experiment 2 sample included 68 children ages 5 and 7, half attended a Montessori school since age 3, half attended a non‐Montessori preschool and one of the three elementary schools (2 public, one private) in the same community	Experiment 1‐ Quasi‐experimental PPCG; Experiment 2‐ Quasi‐experimental PPRM	Experiment 2‐ Three enduring (25 years plus) private Montessori schools. One AMI certified and at others all teachers had AMI or AMS training. Full implementation	Children matched for PPVT. Place value, number line estimation, number ordering, number interpretation, school sale, and multi‐digit calculation	Not specified	The research was funded by a grant from the Institute of Education Sciences
Prendergast 1969	43 children in a Montessori, 41 conventional, and 42 children who did not attend either. All children were from upper‐middle‐class families	Quasi‐experimental PPCG	The intervention lasted 7 months. The Montessori schools were all members of a local association. “The children's experiences followed as closely as possible the activities developed by Maria Montessori.” Medium implementation	Eye‐hand coordination, Figure‐ground, Position in space, Auditory discrimination, Receptive language	1966–1967	None specified
Rathunde 2005a	Montessori (150 students, 60% female, 40% male, 72.6% white, 10.2% Asian, 12.7% Black, 1.9% Latino, 2.6% other races, majority of students from suburban schools in middle‐ or upper‐middle‐class communities). Traditional middle school students (N = 400, half of the students from “ethnic minority families,” matched a subset of this group that was mostly white and middle‐ or upper‐middle‐class to the Montessori group, n = 160, 55% female, 45% male, 74.9% White, 7.8% Asian, 3.6% Latino, and 1.2% other race)	Quasi‐experimental PPRM	Five Montessori schools in four states, selected with consultation with the North American Montessori Teachers Association. AMI teachers. Full implementation	Student Perceptions of their Schools and Teachers (support, order, safety, fairness), Time Use at School, Classroom Activities (passive listening, collaborative work, individual work, media), Time with Friends, Classmates, Teachers, and Alone, Classmates and Friends	Unspecified, but data collection at the traditional schools occurred several years before data collection at the Montessori schools	The O'Shaughnessey Foundation, Dekko Foundation, and Hershey Foundation
Rathunde 2005b	See Rathunde 2005a	Quasi‐experimental PPRM	See Rathunde above	Affect (happy, relaxed, sociable, proud), Potency (strong, active, excited), Intrinsic Motivation (e.g., Did you enjoy what you were doing?, Salience (e.g., Was this activity important to you?)	Unspecified, but data collection at the traditional schools occurred several years before data collection at the Montessori schools	The O'Shaughnessey Foundation, Dekko Foundation, and Hershey Foundation
Tobin 2015	66 refugee children (ages 3–6) in two classrooms, one traditional Thai teacher‐directed (n = 27) and one Montessori (n = 29)	Quasi‐Experimental PPCG	The intervention lasted 54 months. Both Montessori and traditional classrooms have 60 children, one teacher, and one assistant. Unknown implementation	54‐month Ages and Stages Questionnaire: Fine and Gross Motor, Communication, Problem Solving, Personal‐Social	Not specified	None specified
Yussen 1980	Experiment 1: N = 60 4–5‐year‐olds. Half attended Montessori, half attended other private nursery schools. Half boys in each group. Middle and upper‐middle‐class families. Madison, WI	Quasi‐Experimental PPRM	The Montessori schools were AMS or AMI certified. Full implementation	Experiment 1‐ Communication was measured by social cues in different trials, mean length of utterance (MLU) and syntactic complexity of utterances. Emotional perspective–taking measured with tasks to identify and explain different emotions. Peabody Picture Vocabulary pretest for equivalence	Not specified	A grant was received by the Wisconsin Research and Development Center for Cognitive Learning which is further funded by the National Institute of Education. Financial assistance was given by the Spencer Foundation

Note: None of the included studies had disclosed any declarations of interest.

Abbreviations: PCG, posttest‐only with control group design; PPCG, pretest–posttest with control group design; PPRM, pretest–posttest with matched controls.