<![CDATA[This study aims to investigate the implementation of learning using a Problem-Based Learning (PBL) approach based on deep learning to improve students’ mathematical reflective thinking skills. The population of the study consisted of all eighth-grade students of a junior high school in the odd semester of the 2024/2025 academic year, with one class selected as the research subject. The sample was taken using a purposive sampling technique. This study employed a quantitative approach with a quasi-experimental design of the One-Group Pretest–Posttest type. The data were collected through tests of students’ mathematical reflective thinking skills and classroom observations, and analyzed using descriptive and inferential statistics. The results showed that the average pretest score was 58.96, which increased to 74.96 in the posttest. Classroom observations indicated that students actively engaged in problem identification, group discussions, evaluation of solution strategies, and reflective reasoning during the learning process. The Kolmogorov–Smirnov normality test indicated that the data were normally distributed (p = 0.20), and Levene’s test showed homogeneous variances (p = 0.20). The paired-sample t-test revealed a significant improvement in students’ reflective thinking skills (t = -59.084; p < 0.05). The Normalized Gain (N-Gain) calculation showed an average of 0.40, categorized as medium. Thus, the implementation of PBL based on deep learning can significantly enhance students’ mathematical reflective thinking skills.]]>
