<![CDATA[Students’ mathematical problem-solving ability remains relatively low, particularly when dealing with contextual and non-routine mathematical problems. This condition highlights the need for learning environments that encourage collaborative problem solving, meaningful learning, and active student engagement. Therefore, this study aimed to examine the effectiveness of the Student Teams Achievement Division (STAD) learning model with a deep learning approach assisted by Zep Quiz on students’ mathematical problem-solving ability. This study employed a quantitative approach using a quasi-experimental method with a nonequivalent control group design. The research was conducted at Junior High School 1 Paron involving eighth-grade students from an experimental class and a control class selected through cluster random sampling. Data were collected using essay-based pretest and posttest instruments developed based on Polya’s problem-solving indicators. The data were analyzed using normality and homogeneity tests, an independent-samples t-test, and N-Gain analysis. The independent-samples t-test yielded t(58) = 1.589 and p = 0.117, indicating no statistically significant difference between the experimental and control classes. Descriptively, the experimental class showed greater improvement than the control class, as indicated by the higher N-Gain score (0.48 compared with 0.32), although both were categorized as moderate improvement. The effectiveness ratio was 1.52, indicating a descriptive advantage of the experimental class over the control class. This study addresses a gap in previous research by operationally integrating STAD, a deep learning approach emphasizing mindful, meaningful, and joyful learning, and Zep Quiz as an interactive gamification-based medium within a collaborative mathematical problem-solving sequence. This integration connects peer collaboration, meaningful engagement, and interactive digital practice in supporting students’ engagement with contextual and non-routine mathematical problems. Although the difference between the experimental and control classes was not statistically significant, the descriptive findings indicate a positive tendency associated with the implementation of STAD with a deep learning approach assisted by Zep Quiz.]]>
