<![CDATA[The development of coding and artificial intelligence (AI) learning in Indonesia faces the challenge of digital infrastructure inequality, especially in 3T (underdeveloped, frontier, outermost) areas. This study aims to design and test the Box Ocean Ecosystem Sorting Media (BOESM) learning media, an unplugged media that teaches programming logic and AI concepts through marine ecosystem concept classification activities. The study employs a 4D development model (Define, Design, Develop, Disseminate) involving elementary school teachers and students who lack adequate access to digital devices. Data collection techniques include field observation, interviews, questionnaires, expert validation, and analysis of pretests and posttests. BOESM utilizes concrete objects to simulate algorithmic thinking patterns (IF-THEN logic) by grouping marine components, including fish, sand, and plastic waste. The results of expert validation showed that the media was very feasible, and the field trial produced a moderate gain score (0.52), indicating an increase in students' understanding of computational thinking. As many as 92% of students reported enjoying learning without the use of digital devices. The study concluded that BOESM is effective in fostering algorithmic thinking and early AI literacy in technologically constrained environments and is an inclusive and cost-effective learning alternative. ]]>
Copyrights © 2025
