<![CDATA[Purpose: This study examines whether ethnomathematics-based Project-Based Deep Learning (PBDL) is associated with differences in elementary students’ mathematical creativity. Here, deep learning refers to a pedagogical orientation emphasizing meaning-making, reflection, conceptual connections, and durable understanding rather than computational deep learning. Method: A quasi-experimental non-equivalent control-group design involved 60 fifth-grade students from two public elementary schools in Bima Regency, with 30 students in the experimental group and 30 in the control group. Each condition was implemented in a different school, so condition effects could not be separated from possible school- and teacher-level differences. Mathematical creativity was assessed using an open-ended figural-mathematical task covering fluency, flexibility, originality, and elaboration, each weighted at 25 points. Data were analyzed using assumption checks and independent-samples t-tests with mean differences, 95% confidence intervals, and Cohen’s d. Findings: The experimental group had a higher posttest total score than the control group (M = 72.17, SD = 9.84 vs. M = 69.03, SD = 9.25), but the difference was not statistically significant, t(58) = 1.271, p = .209, d = 0.33, 95% CI [-1.80, 8.07]. No creativity indicator differed significantly between groups. Because the experimental group also had a baseline advantage (d = 0.42), the posttest difference should not be interpreted as an intervention effect. Instrument psychometric properties were not fully established, so findings should be interpreted cautiously. Significance: The study documents a six-phase PBDL syntax and identifies requirements for stronger future trials: longer exposure, linked pretest-posttest records, non-confounded allocation, and fully reported scoring and reliability procedures.]]>
Copyrights © 2026
