Evaluasi program implementasi pendekatan deep learning di sekolah menengah pertama

Penelitian

Abstract

This study aims to evaluate the effectiveness of the deep learning approach implementation program at Mesuji 1 State Junior High School in November 2025 using the CIPP (Context, Input, Process, Product) model to support the 21st century learning transformation. The deep learning approach is defined as an in-depth learning methodology that integrates exploration, conceptual understanding, and critical application based on artificial intelligence, aligned with the Independent Curriculum and the 8 graduate profiles. This research was descriptive qualitative with purposive sampling of the principal and two teachers, collecting data through semi-structured interviews and observations. The evaluation showed that the program context was highly relevant to national education policy (100% of respondents); input achieved 70% adequacy with effective teacher training but was hampered by ICT infrastructure; the implementation process was optimal through problem-based learning (PBL) and student-centered simulations despite time constraints and student heterogeneity; and the product resulted in a 15-20% increase in learning achievement, student motivation, and graduate profile achievement, albeit partially due to limited training. This program is innovative in advancing digital education, but requires holistic investment for sustainability, with theoretical contributions to the evaluation of deep learning at the junior high school level as well as practical implications as a reference for replication.