•  
  •  
 

Abstract

Coding learning in elementary schools often emphasizes the implementation of program commands without providing adequate guidance to students to connect conceptual knowledge, algorithmic reasoning, and coding practice. This study aims to develop and evaluate the Knowledge Decoding Learning (KDL) Learning Model integrated with coding and Artificial Intelligence for fifth-grade elementary school students. The study used a research and development method with the ADDIE framework and involved 27 students from SDN Merjosari 1 and 20 students from SDN Merjosari 4, Malang City. Product quality was tested through expert validation, teacher and student response questionnaires, and pretest and posttest analysis using N-Gain. The results showed that the KDL Model obtained a validity percentage of 91% from learning model experts and 89% from coding or digital learning media experts, both of which are included in the very valid category. In large-scale trials, practicality based on teacher responses reached 90% at SDN Merjosari 1 and 84% at SDN Merjosari 4. Ease of understanding based on teacher responses reached 91% and 85%, while student responses reached 89.30% and 83.10%. The average N-Gain at SDN Merjosari 1 was 0.8023 or 80.2280% with an effective category, while at SDN Merjosari 4 it was 0.6218 or 62.1764% with a fairly effective category. These findings indicate that the KDL Model is very valid, practical, easy to understand, and able to improve learning outcomes in schools with different levels of coding learning readiness.

First Page

541

Last Page

559

Download

Share

COinS