Exploration of implementation practices of Montessori education in mainland China

China has prioritized early childhood education (ECE), and the Montessori method has grown widely since the late 1990s, yet little is known about implementation fidelity in Chinese preschools. International research links high-fidelity Montessori implementation to better outcomes in academics, executive function, social development, and motivation. This exploratory study surveyed 210 in-service Montessori teachers and administrators across mainland China to assess classroom practices regarding mixed-age groupings, teaching models (co-teaching vs. single head teacher with assistant), student–teacher ratios, and work cycle lengths. The study addresses two questions: (1) To what extent do early childhood Montessori programs in mainland China reflect high-fidelity practices in these areas? (2) To what extent are departures from fidelity linked to localization pressures within Chinese cultural and regulatory contexts? The topic is important as China may have a very large number of Montessori schools, and fidelity is associated with child outcomes.

Background literature traces Montessori education’s origins and core principles (e.g., normalization through self-directed work) and its global spread, including China’s renewed adoption since the 1990s alongside privatization and local material production. Chinese academic discourse recognizes Montessori’s value but debates its fit within Chinese cultural and policy contexts, advocating for localization to align with Ministry of Education (MOE) guidelines and collectivist cultural values. High-fidelity Montessori indicators identified by organizations like NCMPS include mixed-aged classrooms, a single trained head teacher with a non-teaching assistant, classrooms of no fewer than 24 students with two adults, and uninterrupted work cycles (three hours morning; two hours afternoon for older children). Literature on specific elements: (1) Mixed-aged classrooms are theorized by Montessori to foster social harmony and peer learning; Chinese sources note parental concerns about bullying or academic delay, with some schools reportedly using single-age groupings to appease parents, which would reduce fidelity. (2) Teacher-assistant model: Montessori prescribes one head teacher and a non-teaching assistant; Chinese ECE traditionally employs multiple teachers and emphasizes teacher-directed instruction, potentially pressuring Montessori classrooms toward co-teaching and increased teacher–child interaction. (3) Student–teacher ratios: Montessori supports larger class sizes to promote independence and peer learning, contrasting with typical assumptions favoring small ratios; Chinese guidelines and teacher capacity concerns may drive lower ratios. (4) Work cycles: Montessori emphasizes extended, uninterrupted morning and afternoon cycles to support concentration and normalization; Chinese localization literature suggests more whole-group lessons to foster collectivist identity and alignment with MOE subject-area guidelines, often leading to shortened work periods or omitted afternoon cycles. Overall, the literature anticipates localization-driven adaptations that may reduce fidelity.

Design: Descriptive, exploratory survey of implementation practices in Montessori ECE in mainland China. Instrument: Researcher-developed questionnaire informed by fidelity guidelines (e.g., NCMPS, Lillard) and Chinese localization literature, plus Montessori’s writings. Items covered mixed-age groupings, co-teaching vs. single head teacher, student–teacher ratios, and morning/afternoon work cycle lengths; response formats included multiple-choice, yes/no, and open-ended prompts. Distribution: Via WeChat groups (QR code invitation); respondents completed the survey electronically; single completion enforced. Participants: 210 valid responses—147 classroom teachers (29 with 18 months–3 years; 118 with 3–6 years) and 63 administrators; 93.8% female; 53.8% held bachelor’s or higher (above national preschool average); respondents spanned first-tier to smaller cities nationwide. Analysis: Descriptive statistics; Pearson correlations among morning work cycle length, classroom size, and number of teachers; point-biserial correlation between co-teaching and classroom size; chi-square tests between co-teaching and whole-class lessons, morning circle, and afternoon work cycle; and between whole-class lessons and morning circle and afternoon work cycle. Analyses conducted in SPSS 22. Note: For ratio and work-cycle analyses, 0–3 classrooms were excluded given different developmental norms.

• Sample: 210 respondents (147 teachers; 63 administrators); 93.8% female; 53.8% with bachelor’s or higher.
• Mixed-age classrooms: 96.7% reported mixed-age groupings (15–18 months–3 years; 3–6 years), reflecting high fidelity on this dimension.
• Co-teaching: 87.1% reported more than one teacher allowed to give lessons (co-teaching); 12.9% reported only one teacher giving lessons.
• Classroom size and staffing (3–6 classrooms): Most common sizes were 20 (18.8%) and 25 (29.3%). Grouped categories: 50% had 24 or fewer; 50% had 25 or more students. Among both groups, around three or more teachers were typical (24 or fewer: 75.6% had 3 teachers; 25 or more: 77.8% had 3 teachers). Mode ratio was 25 students with 3 teachers (~8:1); ratios ranged from 15:4 to 50:4.
• Co-teaching by size: 85.6% of classrooms with ≤24 students and 90.0% with ≥25 students practiced co-teaching.
• Work cycles (3–6 classrooms): Morning work cycle—1–2 hours in 66.9%; 2.5–3 hours in 31.5%; 1.6% reported none. Afternoon work cycle—present in 25.2%; absent in 74.8%.
• Group activities: 86% held a morning circle time; 56% gave whole-class lessons in the morning.
• Correlations (n≈210): • Morning work cycle length vs. number of students: r=0.196, p=0.004 (weak positive; larger classes associated with longer work cycles). • Classroom size vs. number of teachers: r=0.225, p=0.001 (weak positive; larger classes had more teachers). • Morning work cycle length vs. number of teachers: no significant correlation. • Co-teaching vs. classroom size (point-biserial): r=−0.094, p=0.176 (not significant).
• Chi-square tests: • Co-teaching vs. whole-class lessons: x²(1)=0.317, p=0.574 (ns). • Co-teaching vs. morning circle: x²(1)=0.255, p=0.774 (ns). • Co-teaching vs. afternoon work cycle: x²(1)=0.317, p=0.636 (ns). • Whole-class lessons vs. morning circle: x²(1)=0.080, p=0.777 (ns). • Whole-class lessons vs. afternoon work cycle: x²(1)=7.341, p=0.007; φ=0.187 (significant minimal association; classes with whole-class lessons more likely to report an afternoon work cycle).

Findings indicate partial adherence to high-fidelity Montessori practices in mainland China: while mixed-age groupings are widely implemented, co-teaching is prevalent, student–teacher ratios are relatively low (more adults per child), morning work cycles are often shortened, and afternoon work cycles are largely omitted. These deviations align with localization pressures and traditional Chinese ECE norms emphasizing teacher-directed, group-based instruction and staffing patterns with multiple teachers. Co-teaching and low ratios may undermine children’s independence and peer-mediated learning central to Montessori’s design. Shortened and interrupted work cycles jeopardize normalization by limiting sustained concentration and self-directed engagement. The study also suggests that larger classes tend to have longer work cycles and more teachers, consistent with regulatory and practical considerations, but co-teaching does not vary by class size, implying a generalized cultural or institutional practice rather than a response to classroom scale. The association between whole-class lessons and the presence of an afternoon work cycle raises questions about the authenticity of these afternoon periods (i.e., independent work vs. additional group instruction). Overall, the departures from fidelity may reduce the benefits documented in high-fidelity Montessori programs, including social cohesion developed through child-led interactions rather than whole-group instruction.

This study provides the first empirical snapshot of Montessori ECE implementation in mainland China relative to high-fidelity standards. High adoption of mixed-age classrooms suggests alignment with Montessori principles; however, widespread co-teaching, lower student–teacher ratios, shortened morning work cycles, and the frequent absence of afternoon work cycles reflect localization and traditional ECE influences. Given evidence that higher fidelity is associated with superior child outcomes, the findings call for reflection among Chinese Montessori educators to reduce teacher-centered practices, ensure a single head teacher with a non-teaching assistant, maintain larger class sizes consistent with Montessori guidance, and protect uninterrupted morning and afternoon work cycles. The study recommends sector-level support, such as a national Montessori alliance, to promote shared standards and professional development focused on authentic implementation, while thoughtfully engaging with cultural and regulatory contexts.

• Generalizability: The convenience sample of 210 respondents, though geographically diverse, cannot represent all Montessori practitioners in China; larger samples are needed.
• Scope of measurement: The study did not assess the presence or use of Montessori materials, nor conduct classroom observations or interviews; future work should include materials audits and qualitative inquiry into teachers’ philosophical understanding.
• Outcomes: No child outcome data were collected; given the link between fidelity and outcomes, longitudinal or comparative studies on student outcomes in Chinese Montessori settings are warranted.