Studying students in Montessori schools.

IN THEIR EDUCATION FORUM “EVALUATING MONTESSORI EDUCATION” (29 SEPT. 2006, P. 1893), A. Lillard and N. Else-Quest present Montessori education as being superior to other types of schools and having “remarkable outcomes.” Unfortunately, the analyses backing these value judgments are plagued by methodological and statistical problems and thus do not provide support for such claims. To evaluate Montessori education, Lillard and Else-Quest measured the performance of children from a single Montessori school—a textbook example of pseudo-replication. The only test performed was thus if this particular school is a good school or not. But not even on that level was the comparison entirely valid, because the pupils in the school were compared with children mainly attending inner-city public schools (72% in 5-year-olds; 78% in 12-yearolds)—the category of schools most often subject to social and economic problems. Further, the sex ratio in the Montessori sample was equal or skewed toward girls (50% girls in 5year-olds; 59% girls in 12-year-olds), while the sex ratio in the inner-city public-school sample was skewed toward boys (60% boys in 5-year-olds; 64% boys in 12-year-olds). Girls in these age groups generally outscore boys (1, 2). However, not even with the study set up in this way did the Montessori school come out on top. Pupils in the girl-dominated Montessori sample rather unsurprisingly scored higher on tests measuring social ability than pupils in the boy-dominated inner-city school sample. But the study also reported that although 5-year-olds in the Montessori sample did better in tests measuring cognitive/academic measures, no such difference could be found in the 12-year-olds. The only really noteworthy result in this study is that for a sample consisting of mainly girls in a particular school in Milwaukee, several years of Montessori education has resulted in no detectable difference in academic performance in comparison with a group of mainly boys having been subject to schooling in Milwaukee inner-city public schools. Because alternative schools can be assumed to attract more dedicated teachers and often have superior resources to public schools (Montessori schools require a high teacher density and a special set of educational materials) and because young girls tend to outscore young boys academically, this is more a quality statement for Milwaukee public schools than anything else. To make informed choices about schooling, parents and policy-makers everywhere are in dire need of proper comparisons between different education systems. Unintended misinformation through poorly performed studies only serves to make the current state of confusion over the pros and cons of various education systems worse. PATRIK LINDENFORS