Development and validation of a deep learning-based physical education e-module for fundamental movement skills

Abstract

Purpose: This study aimed to develop and determine the feasibility and practicality of a Deep Learning-based Physical Education e-module for teaching fundamental movement skills to Grade IV elementary-school students. Method: A Research and Development approach using the ADDIE model was employed. Data were collected through needs analysis, expert validation, and practicality questionnaires and analyzed using descriptive statistics, Aiken’s V, item validity, and Cronbach’s alpha. Results: Expert validation demonstrated good content (84%; V = 0.795) and media feasibility (83%; V = 0.783) and very good language feasibility (97%; V = 0.958). The practicality instruments showed high reliability (α = 0.862–0.876), while teacher/practitioner assessment indicated very high practicality (92.50%). Conclusion: The developed e-module was feasible, reliable, and highly practical for supporting fundamental movement instruction through mindful, meaningful, and joyful learning in elementary Physical Education