In this preregistered systematic review, we identified 198 studies, of which 114 provided sufficient quantitative data for meta-analysis. Across these studies, the Montessori method was associated with nominally positive and conventionally large, pooled effects on learning outcomes in pretest–posttest and posttest-only designs. However, these estimates were characterized by extreme between-study heterogeneity, substantial small-study effects, and clear indications of publication bias, which substantially limited their robustness and interpretability. Methodological limitations, including inconsistent reporting of implementation fidelity, low transparency, and limited experimental control, further constrained the generalizability of the findings. By combining design-sensitive modeling, multimethod bias diagnostics, and an operational framework for Montessori fidelity, this review advances a metascientific perspective on the credibility of evidence in education research. Overall, the results indicate that the large average effects reported in the literature are largely attributable to structural weaknesses of the evidence base rather than to robust educational impacts, underscoring the need for greater methodological rigor, preregistration, and open data to strengthen cumulative knowledge about Montessori education.
León et al. (Tue,) studied this question.