Cultivating mathematical problem-solving skills in early education is foundational to fostering logical reasoning and analytical thinking. Traditional mathematics instruction often relies on rote memorization, leading to low student engagement and weak conceptual application, while empirical evidence evaluating the structural impact of integrated Science, Technology, Engineering, and Mathematics (STEM) models in primary mathematics remains critically limited. This study aims to examine the effect of the STEM education model on elementary school students' problem-solving abilities in mathematics learning. Utilizing a quantitative quasi-experimental framework, this study selected students from SD Negeri Tanjungsari and SD Negeri Kadukacapi as research samples, systematically categorized into experimental (STEM-based instruction) and control (conventional instruction) groups. Quantitative data were gathered via a validated mathematical problem-solving test and analyzed using descriptive metrics (mean, standard deviation, and learning completeness percentages) alongside paired and independent sample t-tests. The empirical findings indicate that the STEM education model exerts a profoundly positive and statistically significant effect on students' mathematical problem-solving abilities. The experimental class achieved a substantially higher mean score (93.92) compared to the control class (61.34). Furthermore, a narrower standard deviation in the experimental cohort indicated a more equitable distribution of skill mastery. The student learning completeness percentage also demonstrated a staggering disparity, with the experimental class reaching 94.33% mastery against the control group's mere 71.66% (p < 0.05). Consequently, the STEM education model serves as a potent pedagogical alternative to address the pervasive deficit in elementary students' mathematical problem-solving. Beyond individual performance metrics, these insights offer a robust framework for educational authorities to design interdisciplinary curricula that effectively bridge abstract mathematical principles with real-world, innovative applications.
Copyrights © 2026