DEVELOPMENT OF A DEEP LEARNING MODEL: CONTEXTUAL STRATEGIES FOR IMPROVING THE PROFILE OF PANCASILA HIGH SCHOOL STUDENTS

A deep learning-based learning model was developed to enhance students' reflective and contextual understanding of learning materials while internalizing the values of the Pancasila Student Profile. This study aimed to create a contextual learning model for high school students in science class, employing a quasi-experimental design with a non-equivalent control group. Data were collected via questionnaires and analyzed using descriptive statistics and inferential tests, including the Independent Samples t-Test and Paired Samples t-Test, to evaluate the model's effectiveness in improving student learning outcomes. The results revealed a significant difference in the experimental group, as the Paired Samples t-Test indicated a notable difference between the pre-test score (68.47) and the post-test score (83.20), with a t-value of -9.127 and p = 0.000 (p < 0.05). Additionally, the Independent Samples t-Test demonstrated that the average post-test score of the experimental group (83.20) was significantly higher than that of the control group (74.13), with a t-value of 4.213 and p = 0.000 (p < 0.05). These findings indicate that the local wisdom-based model effectively improves student learning outcomes compared to the conventional approach. This study highlights the importance of implementing an immersive learning model in high schools, emphasizing contextual strategies that engage students in critical and reflective thinking to internalize the values of the Pancasila Student Proficiency Profile. However, the limitations of this model, such as time and space constraints, suggest the need for further longitudinal research across various educational settings to comprehensively evaluate its effectiveness and adaptability.