Differentiated Instruction Based on Deep Learning in Science Education for Grade XI MDVI Students

Abstract

This study examines the implementation of differentiated instruction based on Deep Learning within the Merdeka Curriculum in science instruction for Grade XI students with MDVI at a special senior secondary school, with the objectives of describing their learning capabilities, identifying the difficulties the teacher faces, analysing the efforts undertaken to overcome those difficulties, and examining how the differentiation is enacted. The study employed a descriptive qualitative case study within an interpretive paradigm. Data were collected through semi-structured interviews with the science teacher, non-participatory classroom observation, and documentary study of initial assessment records, the Individualised Education Programme, worksheets, and student portfolios, and were analysed using the interactive model of Miles, Huberman, and Saldaña; data validity was maintained through source and technique triangulation, prolonged observation, and thick description. The findings reveal sharply contrasting capability profiles: one student accesses material through Braille and tactile exploration, while the other relies on auditory-verbal processing. The teacher encounters five interrelated difficulties, namely concretising abstract concepts, linking material to students' interests, sustaining students' mood, serving differences in readiness simultaneously, and coping with limited adaptive media and distraction in a shared classroom, and responds through concrete media drawn from the immediate environment, an individualised approach grounded in initial assessment, the use of students' interests as an entry point, and the connection of material to real experience. Differentiation of content, process, and product is enacted systematically, while deep learning is realised through connection to everyday life and routine reflective evaluation.