<![CDATA[While deep learning has become a focal point in the transformation of Indonesian education, its implementation requires further investigation. This study aimed to measure the effectiveness of a deep learning approach in enhancing AI literacy among elementary school students through the implementation of the AMAIL (Associative Model of AI Literacy) model. The study employed a pre-test post-test design across five intervention cycles, involving 118 fifth-grade students from three elementary schools in Salatiga. Data analysis utilized bootstrap techniques for Likert-scale questions and Exact McNemar's tests for dichotomous questions. Results showed a significant improvement in understanding AI personalization (p=0.020) and consistent growth in students' ability to explain AI concepts and demonstrate ethical awareness regarding AI use in later cycles (p<0.001). The developmental trajectory from the first to the fifth cycle underscores the importance of the implementation refinement process for the success of deep learning in enhancing students' AI literacy. This research provides an empirical foundation for developing AI literacy programs using a deep learning approach for elementary school students in Indonesia.]]>
Copyrights © 2025
