Deep Learning Training: Improving Teacher Competence in Implementing “the 2024 Merdeka Curriculum”

Abstract

The transition to a competency-based curriculum requires a paradigm shift for teachers, from mere implementers of material to innovative curriculum developers. However, teacher training is often theoretical and fails to address substantive classroom practice. This study aims to document and analyze the effectiveness of Deep Learning strategy-based training in improving the competency of high school teachers in DKI Jakarta to authentically implement the 2024 Independent Curriculum. Using the ADDIE (Analysis, Design, Development, Implementation, Evaluation) instructional design framework, this activity involved 224 teachers in the online phase and 33 selected teachers per class in the offline phase. Data were collected through needs surveys, observations of the results of the Learning Objective Flow (ATP) development, and pre- and post-tests related to pedagogical strategies. The results showed an increase in participants' knowledge outcomes from an average pretest of 68.5 to an average posttest of 83.7 with an N-gain of 0.78. The Deep Learning approach significantly improves teachers' pedagogical literacy, as evidenced by their ability to deconstruct Learning Outcomes (LO) into contextual instructional designs based on deep conceptual understanding. The transformation in mindset is evident in the shift from conventional teaching methods to reflective and inclusive learning strategies. This study concludes that training that emphasizes independent knowledge construction and collaboration is the main key in bridging the gap between macro curriculum policies and micropedagogical practices in the classroom.