This was written by Tim Seldin and published in Tomorrow's Child in Spring, 2003. In the last several years, a great deal of work has been completed, some of which has been surveyed above. There is widespread agreement that more studies need to be completed by researchers who are not connected to the Montessori community, and that focus needs to be on Montessori in the public sector so that parent ability to pay tuition is not a factor that calls data into question.
As Montessori expands into the elementary and secondary levels, Montessori educators are increasingly asked for evidence that Montessori works.
More than 200 studies have been done about the long-term effects on children who have attended Montessori schools Montessori in the United States since Montessori's resurgence in this country in 1960. However, the research that has been done to date on Montessori's effectiveness in the American educational context is far more limited than it should have been after its more than 90 year history in the United States.
As Dr. John Chattin-McNichols points out, “there is widespread lack of knowledge about the Montessori method on the part of traditional early childhood educators, and in the mainstream of American education. Montessori supporters have typically not strongly supported research efforts, or the communication of the results of research on Montessori to those outside of the Montessori community. Thus, despite the existence of more than 200 studies on Montessori, neither Montessori adherents nor the mainstream educator is aware of these findings. The problem is made more difficult by virtue of the inaccessibility of many of the articles, since many of them are "fugitive" documents: publications (such as the journals of the Association Montessori Internationale or the American Montessori Society) which are not indexed in ERIC or other abstracting services, conference papers, or unpublished dissertations.”
Another partial explanation for the limited number of studies may be that until the mid-1980s, Montessori programs typically ended at age six. Elementary programs have only become widespread in the1990s, and middle and high school programs are only now beginning to slowly spread. While research into early childhood education has been done for some time, most of America's attention tends to be focused on the elementary and secondary levels, as a result of which, few Montessori educators felt great pressure to gather research data until fairly recently. As a result, even though the Montessori method has been used successfully in schools around the world since the early 1900s (extending up through high school in some countries), in the United States, Montessori has grown without a conscious effort to document its effectiveness through research.
Montessori researchers, Mary Boehnlein and John Chattin-McNichols, have suggested several obstacles that have hindered the implementation of more definitive studies (Public School Montessorian, Volume 2/Number 2, Winter, 1990, p.1):
Shared Definitions Are a Prerequisite for Research. What is an 'Authentic Montessori "classroom? There is no universally accepted definition. It is not clear whether the program called "Montessori" in one study is identical, or even similar, to programs elsewhere.
What is a reasonable time frame for evaluating the effects of a Montessori experience? Can a study of achievement after six months be accepted as a legitimate measure? After one year? How long must a child be in a Montessori classroom before one can assume subsequent changes are directly correlated to the classroom experience?
What must a researcher understand about Montessori education before the research can be judged valid? Boehnlein wrote: "If the researcher of the Montessori method does not understand the philosophical basis of Montessori's 'Cosmic Education,' the research is compromised and the results meaningless."
Can an evaluation that focuses on achievement preserve the integrity of a program that is concerned with a broader definition of educational purpose? "We're going to have to develop instruments of our own to measure things we claim Montessori education does," Boehnlein wrote. "We'll have to look at affective (emotional development) areas and social consciousness. But the public schools respond to achievement. We have to document that. Then we have a job to educate them about what we're really about. I think the public schools would welcome data that helps them show they produce not only high achievers but responsible citizens who care about their environment and their world.”
In general, a review of the complete literature is somewhat frustrating. Many studies yielded inconclusive results or were seriously flawed for reasons such as working with too small a student population to yield a reliable sample, as well as factors that call into question whether the program studied was an authentic and credible example of good Montessori practice. Some were challenged by the difficulty of defining how one measures the development of young children, especially in areas like social skills, behavior, self-motivation, and mental health, all of which would presumably distinguish Montessori children from the general population.
This article attempts to highlight some of the most important studies that have been done to date. It is very limited by the space available in a publication like Tomorrow's Child. The information included was drawn from several excellent summaries, including those prepared by Dr. Mary Boehnlein, Dr. John Chattin-McNichols, Dr. Christopher Glenn, and Dr. Ginger McKensie. For a more complete bibliography, email us at firstname.lastname@example.org.
Some of the Early Studies in Montessori's Effectiveness
Many of the earliest studies into the Montessori approach were done in the 1960s, 1970s and 1980s in conjunction with programs serving disadvantaged minority-group children, many of who were enrolled in Montessori Head Start programs around the United States. More recent studies are beginning to examine the results of Montessori programs in public schools around the country, as well as a few interesting studies that follow students at more affluent independent Montessori schools from their early years into high school, and in at least one case, adulthood.
As the North American Montessori Teachers' Association summarizes their assessment of the results of a number of key research abstracts, “studies such as those by Takacs, Karnes, and Duax show that Montessori education at both the preschool and elementary levels benefits low-SES children by giving them higher competence in basic skills, a better attitude toward school, and a greater chance of staying in school than their non-Montessori peers possess.”
The emphasis on the positive evidence of Montessori's effect on disadvantaged students is especially significant, as it has been suggested over the years that the success of Montessori students who come from middle and upper class backgrounds may be as much connected to the advantages provided by their socioeconomic status as to the benefits Montessori offered them.
In 1976, Sciarra and Dorsey conducted a six-year follow up study of children who participated in an earlier research study. It was designed to evaluate whether or not early and continued exposure to Montessori education makes a difference in later academic achievement. Third grade children who had varied preschool and primary school experiences were evaluated and compared in both verbal and mathematics skills as measured by the Metropolitan Achievement Test.
The children were evaluated as four groups. The first group of children had four years of Montessori education including preschool and elementary school. The second group of children had two years of Montessori preschool. The third group had one year of Head Start prior to Kindergarten, and the fourth group of children had no school experience prior to kindergarten.
The results of the six-year follow-up study demonstrated that the group with four years of Montessori education, including preschool and elementary school, scored best on all seven variables of the third grade level Metropolitan Achievement Test. The purpose of this nine-year follow-up study was to investigate whether these positive effects were maintained up to sixth grade level. The study showed that children who had attended Montessori preschool scored higher on subtests of the Metropolitan Achievement Test administered at sixth grade level than did those children who had had Head Start or no preschool.
Differential Outcomes of a Montessori Curriculum In 1972 Stodolsky and Karlson reported the results of a study of the effects of attendance at a Montessori preschool on a group of disadvantaged children and a group of middle class children who attended the same school and were integrated into the same classrooms. The children were individually tested using the Stanford-Binet Intelligence Scale. All sample children received selected scales from the Wechsler Preschool and Primary Scale of Intelligence. The study demonstrated that the Montessori curriculum is effective in nurturing continuing development in both disadvantaged and more affluent children in the areas of visual-motor integration, matching and sorting skills, psychomotor skills, and number concepts.
Four Preschool Programs: Their Lasting Effects In 1979, Jones and Miller reported the results of a study of the long-term effects of preschool experience on sixth and seventh grade students. The subjects in this study were 200 primarily black, Head Start children, who in 1966 were randomly assigned to one of four preschool programs (Bereiter-Engelmann, Decree, Montessori, and Traditional), In 1976-77, the children, now in seventh grade, were given the Wechsler Intelligence Scale for Children-revised and the Stanford Achievement Test. Results indicated that the performance of the students who had attended a Montessori program were consistently higher than those who had attended one of the other three programs.
The researchers noted that “Several hypothesis can be offered for the consistent superiority of Montessori children over those from the other three programs from first grade on. The Montessori program views the young child as naturally curious, innately eager to learn, and capable of intense concentration. Perhaps this attitude on the part of the teachers became a self-fulfilling prophecy. We did not measure self-esteem. Possibly the Montessori children came to see themselves more as ‘students.'
Montessori and Responsive Environment Models: A Longitudinal Study of Two Preschool Programs In 1977 Seefeld completed a study that was a continuation of a longitudinal assessment of Montessori and a second model of preschool education (Responsive Environment) programs sponsored by the Arlington County, Virginia Public Schools. The Metropolitan Readiness Test, the Caldwell Cooperative Preschool Inventory and the Peabody Picture Vocabulary Test were used to assess the academic achievement and intellectual development of 3, 4, and 5-year old children who attended either a Montessori or Responsive Environment preschool program, along with those who had no preschool experience, at the end of the regular kindergarten program. The SRA Achievement Series, Grade 1, was used to assess the achievement of children, with and without the Montessori experience, at the end of first grade. The results indicated that children in the regular five-year old kindergarten program with prior Montessori experience scored significantly higher than did children without Montessori experience prior to entrance into the elementary program.
In 1969 Di Lorenzo studied the effects of a year prekindergarten school experience on l,807 low-income children. Again, the majority of the children studied were Black. The study was a true experimental design, with stratified random assignment, in eight school districts in New York. Four of the programs were described as early childhood oriented, four as cognitively oriented. Of the four cognitively oriented programs, one was a Montessori program. Three waves of children were tested on measures including the Stanford-Binet IQ test, the Peabody Picture Vocabulary Test, the Illinois Test of Psycholinguistic Abilities, the Learner Self-Concept Test, and the Metropolitan Readiness Test, at the beginning and end of prekindergarten, kindergarten, and first grade years.
Post-tests at the end of one year and follow-up testing at the end of kindergarten indicated that pupils in the four cognitively oriented programs, one of which was a Montessori program, surpassed those in the four traditional nursery school programs on Stanford-Binet and Peabody Picture Vocabulary Test scores. The Montessori program was found to be associated with modest but significant gains in Stanford-Binet IQ.
The follow-up study by Miller and Bizzell, (1983,1985) Miller and Bizzell elaborately analyzed the dropout patterns among the groups and other possible spurious causes of this result, and concluded that the Montessori program itself was responsible for the achievement test superiority.
Long-Term Effects of Four Preschool Programs: Sixth, Seventh, and Eighth Grades In 1983, Miller and Bizzell conducted a follow-up study that showed that those children who had attended Montessori programs began to display clearly superior school achievement both in reading and math among by the end of second grade and were still present at the end of sixth grade.
Fleege, Black, and Rackauskas (l967) compared children (N = l24) in public and private school from Montessori preschool, non-Montessori preschool, and no-preschool background. Teachers rated the children in several areas on an experimental questionnaire. Montessori children were rated significantly superior on interest in learning, independence, interpersonal relations, leadership, and learning ability.
Kohlberg (l968) reported finding that in a yearlong Montessori program, ten low-income Black children studied showed a mean Stanford-Binet IQ increase of l7 points between October and January. The children of average IQ on the first testing increased as much as the children of low IQ. A group of middle-income children in the same class showed a mean increase of l0 IQ points.
In a related study, Pendergast (l969) administered the Frostig Developmental Tests of Visual Perception twice, seven months apart, to the same populations in order to evaluate eye-hand coordination and visual perception skills. The results found that the Montessori children showed significantly greater gains in eye-hand coordination than those who attended the conventional nursery school.
Long-range studies by Karnes (1969, 1978, 1983), found that after just one year of Montessori preschool, low socio-economic status children showed "superior performance on measures of autonomy and curiosity" over children from similar socio-economic backgrounds who attended other preschool programs. Karnes also found higher school success ratings, and the highest percentage of high school graduates among her Montessori group. Studies found that although disadvantaged children from Montessori preschools showed that in later years, low-income students who attended Montessori programs exceeded their peers who did not have a Montessori experience in academic competence and achievement as well as attitude toward school. Moreover, data showed that significantly higher numbers of Montessori children completed high school.
A long term follow up of Karnes' research published in 1985 is important, since it is, along with the Miller and Dyer study below, one of the few long term studies of the effects of Montessori programs. Overall, in a composite measure of the effects of four preschool programs on a wide range of measures, the Montessori program was rated as the most effective in producing long term school success, ahead of the other behaviorally oriented preschool programs.
Miller and Dyer (l975) studied changes on several measures of general verbal intelligence after one year of Montessori preschool. Stanford-Binet scores averaged across the whole group rose from two to ten points. The average of the two Montessori classes was a larger gain than the DARCEE or control groups' average gain and slightly less than the gains for the Bereiter-Engelmann or traditional nursery school groups at the end of the preschool year.
The Sands School Project (Banta l969) tested children with preschool and primary Montessori experience against control groups with and without a non-Montessori preschool year. The results found the Montessori method to be effective in nurturing development in the areas of visual-motor coordination and integration, matching appropriate objects, and visual-perceptual ability.
Fleege, Black, and Rackauskas (l967) matched a group of 2l children from a Montessori classroom with another 2l from a non-Montessori preschool on variables including age, sex, IQ, birth order, SES, and parental education. The two groups were compared on standardized tests (Peabody Picture Vocabulary Test, SRA Primary Mental Abilities) and on ratings by the teacher in eight areas of competence. A second phase compared achievement test scores and teacher ratings of children in elementary grades who had had Montessori, other preschool, or no-preschool experience. In the comparison between the two matched groups, a complex analysis showed superiority of Montessori children on a verbal ability factor. The comparison of teacher ratings showed "no particular adjustment problems peculiar to Montessori trained children.
Sciarra and Dorsey (1974) followed up the children in the Sands School Project on the Montessori Achievement Test. This is an important study as it represents one of the few reported empirical studies on the effects of Montessori elementary school experience. Because the children ranged in age from eight to almost ten years, age was used as a covariate to adjust the scores. The use of analysis of covariance (see Cronbach et al. l977), the small number of subjects, and the possibility of differential attrition by groups all suggest caution in the interpretation of the results. The authors conclude that continued experience in this Montessori system increases the benefits of preschool exposure. This is interesting in the context of the Miller and Dyer (l975) findings. Perhaps the most important conclusion from this area is that the effects of Montessori elementary school should be investigated in more detail.
Marcella Dawson, in an unpublished Master's thesis (1988), examined the academic performance of public elementary Montessori school students by race. She found that students in Houston's Dodson Montessori Magnet consistently scored above grade level means on standardized tests and moved farther above the mean with each year they remained in the program. This study strongly supported the effectiveness of the Montessori magnet schools in Houston for children of all races, but especially for Hispanics.
Dawson studied student scores on the Iowa Test of Basic Skills and Metropolitan Achievement Test 6. Her sample included 88 of the 125 students enrolled in Dodson's Montessori magnet grades 1-5 in 1987-88. The racial composition was 31 percent Black, 39 percent white, 26 percent Hispanic and 4 percent Asian. Although students entering Dodson are expected to be at the development grade level, there is no cut-off level nor any academic or performance criteria for admission. "The population is representative of the district-wide population," she wrote.
Although scores of Black and Hispanic young people were lower than those of white children in the program, they were significantly higher than scores of Black and Hispanic children in single-race schools. That data, Dawson wrote, suggests that the Montessori program is advantageous for minority students.
Chattin-McNichols points out that Dawson's conclusion is significant because her study is one of the few that tracked achievement in a public Montessori elementary program. "This superior performance," Dawson wrote, "is seen for all subtest topics and at all grade levels in the program. The Montessori advantage of years-in-program produces a steady increase in above grade level achievement, the fifth year indicating the highest gains two years above grade level... The Montessori program at Dodson does not eliminate racial differences [in test scores], but it does leave all races performing above grade level on standardized tests."
Takacs and Clifford (1988) report on a sample of graduates from the Cleveland public Montessori program. In the Reading subtest of the California Achievement Test, the Montessori group had higher scores: 62nd percentile vs. 48th percentile for a matched comparison group. On a teacher survey form, again several significant differences were found in areas such as Pursuit of a Task beyond the Minimum.
In a 1991 study, Takacs, graduates of the Montessori Head Start program at the Marotta Montessori Schools of Cleveland (an inner city program serving low income children) who had entered the Cleveland Public Schools (CPS) were studied in relation to their CPS peers. California Achievement Test scores for Marotta graduates in grades one through eight were compared with the overall scores of first- through eighth-graders in the Cleveland Public Schools. The results showed that the former Montessori students consistently fared significantly better than students who had no previous Montessori education.
In a third study completed in 1993, Takacs found the graduates of Marotta Montessori's Head Start program, far surpassed fellow students who had no Montessori experience in eligibility for the gifted and talented program in the Cleveland Public Schools. Overall, in 1991, only about 4% percent of the population of school aged children in Cleveland qualified for the gifted and talented program by scoring at or above the 75th percentile nationally on the California Achievement Tests, whereas the percentage of the children who had attended Marotta and qualified for the gifted program was dramatically higher: Grade One: 33% , Grade Two: 25%, Grade Three: 20%, Grade Four: 50%, Grade Five: 10%, Grade Six: 43%, Grade Seven: 30%.
In addition to the Takacs work, in 1990 Dr. Mary Maher Boehnlein reviewed 244 studies of Montessori pedagogy, including 25 that focused exclusively on children of low socio-economic status (SES). She found these studies to show overall that "low SES children benefited significantly" from Montessori preschool, even if they attended for less than the full three years.
Other studies confirm these results and point toward even better results for low-SES children who attend Montessori programs throughout the preschool and elementary years.
In 1989 Dr. Tim Duax studied the 1987 and 1988 graduates of MacDowell School, a Milwaukee public-school Montessori program spanning ages 4 to 11. Of these students, 36% were eligible for the federal hot-lunch program because of low parental income, and 50% were minority. His findings were very similar to Dawson's: 84% of the graduates tested above the 50th percentile on the Iowa Test of Basic Skills, reflecting a level of achievement two years in a row substantially higher than national norms. Two other statistical results further illustrate the significance of these results. In comparison, nationally, 23% of all students scored in the "high achievement" range on the Iowa test of Basic Skills, while 44.5% of MacDowell's inner city Montessori program graduates scored in that range. Likewise, while 23% of all students across the nation score in the "low achievement" range, only 1.2% of MacDowell graduates scored in that range. Of the entire two years of graduating classes, only one student did not scores at or above the average score.
In another study completed in 1989, Duax asked 27 middle-school teachers in three Milwaukee public middle schools to assess 15 randomly-selected graduates of MacDowell's prekindergarten and elementary Montessori program in comparison to other students in the same schools who had no Montessori background. (The 15 were chosen to reflect the total graduate population of MacDowell in gender, race, and socioeconomic status factors.) The teachers gave the Montessori-prepared sample above-average ratings in relation to their peers on each of the 25 characteristics on the survey (such as following directions, turning in work on time, listening attentively, using basic skills, asking provocative questions, and adapting to new situations). Duax found that the MacDowell sample ranked highest on 5 characteristics: using basic skills, being responsible, showing enthusiasm for class topics, being individualistic, and exhibiting multicultural awareness.
A Few More Studies
An 18-Year Longitudinal Study of the Success of Montessori Students
Christopher Glenn, Consulting Research Director at Franciscan Montessori Earth School, conducted a number of studies analyzing the student body of this private Montessori school and comparing there development to students in traditional school programs. In 1984, he began an 18-year longitudinal study to assess the effects of a Montessori education well into adulthood.
The study attempted to address two questions often expressed by parents and others:
• Do 12 to 15 years of Montessori education prepare a student for the "real world?
• Do students who continue with their Montessori education after kindergarten tend to develop more "Montessori-like" personalities, interests, and behaviors?
Participants were assessed once every three years (one Montessori cycle). Results supported the studies two primary hypotheses that the number of Montessori Education Years would be positively related to those qualities which are emphasized in the Montessori teaching environment and that children who attend Montessori programs would be successful in later life.
By the spring of 1993, students were found on all assessment criteria to “fall into the range which was best described as normal or healthy social and emotional development. Achievement test results were above national norms on all scales.” The students, both those still attending Montessori and those now attending other schools, were described by their parents to be easy to communicate with, willing to do household chores, and generally positive members of the family. Present teachers rated them as having a high ability to fit in socially, cooperate with teachers and peers, work alone, finish a product, handle stress, and accomplish academic tasks. This was particularly pronounced in the case of Montessori educated students who were now attending traditional schools when compared by their teachers to other students who had not attended Montessori school in earlier grades. Students also reported that they felt positively about their school experience as well.
In 1999 Glenn conducted a13 year follow-up which showed that students continued to display healthy behavior and adjustment and above average academic performance on all scales. Teachers rated the now former Montessori students as performing better and behaving with more maturity than other class members. Students were very positive about the long-term effects of their Montessori experience. The personal value of lifelong learning was identified as most prevalent among the students who continued with Montessori at least through the elementary years. Students were described as preferring direct experiential (hands-on) learning, one-on-one learning (and peer teaching), self-direction and control, and travel. Students were also described as having a strong desire for self-understanding, general personality development, self-direction and discipline, and a strong positive attitude towards social-interactive activities.
This long-term study continues, following the students through the college years into young adulthood.
And finally, a new study being prepared for publication provides some of the most promising results to date on the effects of public Montessori schools. In the Fall, 2002 issue of Public School Montessorian, it was reported that Alan Gartner and Dorothy Lipsky, respected academic researchers at City University of New York announced at last July's conference of the Association Montessori International/USA that “graduates of Milwaukee's public Montessori program performed significantly better than peers on some measures of science, social science and mathematics performance.”
The study compared 201 high school students who completed fifth grade in Milwaukee's two public Montessori schools between 1990 and 1994 to a matched set of peers with similar socio-economic indicators and mobility. The study looked at factors such as GPA, standardized test scores, honors, failed classes, attendance and discipline issues. The study revealed:
• Math and social studies GPAs were higher for Montessori graduates, although there was no significant difference in over-all GPAs.
• On the district's standardized tests, male Montessori graduates did significantly better than their peers in science and female Montessori graduates did better than their peers in math.
• Montessori graduates as a group did significantly better than their peers on the ACT mathematics exam.
• the Montessori graduates took more honors courses and failed fewer courses.
Gartner and Lipsky reported that they were impressed by the enduring effects of the Montessori program. “Little educational research looks at the staying power (of a program),” Gartner said.
“These results, seven years after the intervention, indicate a powerful effect. This is real staying power.”
The research project was initiated by Association Montessori International/USA, which had a major hand in implementing the first two Milwaukee Montessori programs at MacDowell and Greenfield schools.
In conclusion, much more research needs to be done, and the Montessori community needs to do a much better job of disseminating information about existing research and new studies as they come in.
The studies referenced in the summary above are just a sample of the more than 200 studies that have been published to date. For additional information about these and other projects, see the following sources:
Boehnlein, M. Montessori Research: Analysis in Retrospect. Special Edition of the North American Montessori Teachers' Association Journal, Vol. 13, No. 3, Summer, 1988
Boehnlein, M. The NAMTA Montessori Bibliography. Special Edition of the North American Montessori Teachers' Association Journal, Vol. 10, No. 2, Summer, 1985
Boehnlein, M (1990). Research and evaluation summary of Montessori programs. In David Kahn, ed. Implementing Montessori Education in the Public Sector. Cleveland: North American Montessori Teachers' Association. pp. 476-483.
Chattin-McNichols, J. The Effects of Montessori School Experience. Young Children Vol. 36, 1981
Chattin-McNichols, J. and Loeffler, M. Teachers as Researchers: The First Cycle of the Teachers' Research Network Young Children Vol. 44, No. 5, 1989
Chattin-McNichols, J. The Montessori Controversy Albany, NY: Delmar Publishers, in press 1990
Chattin-McNichols, J. What Does Research Say About Montessori? Chapter in Montessori in Contemporary American Society, (Ed. Margaret Loeffler) Albany: Heinemann, 1992
Chattin-McNichols, J. Montessori Programs in Public Schools. ERIC Digest ED0-PS-92-7. 1992
Chattin-McNichols, J. What Does Research Say About Montessori? Paper presented at the November, 2002 conference of the International Association of Montessori Educators
The ERIC Database: The Educational Resources Information Center (ERIC) is a national information system designed to provide access to education-related literature. The ERIC database is the world's largest source of educational information http://ericeece.org
Glenn, C. A Comparison of Lower and Upper Elementary Montessori Students with a Public School Sample. North American Montessori Teachers' Quarterly, Vol. 14, No. 2, 1989
Glenn, C. The Longitudinal Assessment Study: Cycle 3 (Seven Year) Follow Up. ERIC Documents Reproduction Service, No. ED 370679, 1993
Glenn, C. The Longitudinal Assessment Study: Cycle 4 (Ten Year) Follow Up. ERIC Documents Reproduction Service, No. ED 403013, 1996
Glenn, C. The Longitudinal Assessment Study: Thirteen Year Follow Up. ERIC Documents Reproduction Service, No. ED 431543, 1999
McKensie, Ginger Nine Significant Studies Public School Montessorian, Winter, 2002, Volume 2, Number 2, p. 14-15