This study analyzes the impact of implementing a deep learning-based flipped classroom model on Fiqh madrasah students' learning outcomes in madrasah examinations, focusing on developing contextual and meaningful learning. The study uses a quantitative approach with a quasi-experimental pretest-posttest control group design. The research subjects consisted of 70 students divided into a group using the deep learning-based flipped classroom model and a group receiving conventional instruction. Data were collected through learning outcome tests, observation, and documentation. Data analysis used descriptive statistics, the N-Gain test, and the independent-samples t-test, after confirming normality and homogeneity assumptions. The results indicate a significant difference in learning outcomes between the deep learning-based flipped classroom group and the conventional group, with a significance value of 0.031 (<0.05). The experimental group achieved a mean posttest score of 87.65 with an N-Gain of 0.64, whereas the conventional group obtained a mean of 82.67 with an N-Gain of 0.54. The study's implications suggest that the deep learning-based flipped classroom model serves as an alternative instructional strategy for Fiqh that can enhance student engagement, strengthen conceptual understanding, and support the implementation of innovative learning practices in madrasah.
Copyrights © 2026