Implementing Deep Learning-Oriented Academic Supervision in Vocational Schools

Abstract

The declining quality of learning in Vocational High Schools, reflected in the 2025 Education Report, highlights the need for more effective strategies to strengthen instructional practices. This study aims to examine the implementation of deep learning–based academic supervision and to identify the factors influencing its effectiveness at SMKS Jayanegara Lawang, Malang Regency. A qualitative case study design was applied, involving school leaders and teachers as key informants. Data were collected through interviews, observations, and documentation, and analyzed using the interactive model of Miles, Huberman, and Saldana. The findings indicate that academic supervision follows a structured, continuous cycle comprising planning, classroom implementation, evaluation, reflective feedback, and systematic follow-up. This process forms a supervision model that promotes student-centered learning, characterized by meaningful engagement, reflective thinking, and contextual application aligned with workplace demands. Its effectiveness is supported by teacher commitment, collaborative culture, digital readiness, and institutional support, while constraints arise from resistance to pedagogical change, limited infrastructure, time allocation issues, and learner diversity. The study contributes by proposing a practical supervision framework that integrates deep learning principles into professional development processes. It also implies that policymakers and school leaders should strengthen capacity-building programs, improve infrastructure, and institutionalize reflective supervision practices to enhance learning quality in vocational education.