Distinguishing fact from opinion is an essential information literacy skill for elementary students, yet achievement in Indonesia remains below minimum standards because instruction relies predominantly on text-based media. This study developed, validated, and tested the initial effectiveness of an Edpuzzle-based interactive video for fact-and-opinion instruction in Grade 5. Development followed the ADDIE model and applied principles of generative processing from the Cognitive Theory of Multimedia Learning through embedded questions and formative feedback. Three experts assessed content validity using Aiken’s V. Instrument quality was evaluated through item validity, KR-20 reliability, difficulty index, and discrimination index. Effectiveness was tested using a one-group pretest–posttest design (n = 35) with a paired-samples t-test, Cohen’s d, and normalized gain (N-Gain). Content validity yielded Aiken's V = 0.793, KR-20 reliability = 0.880, and a student response rate of 90.89%. Posttest scores (M = 83.83) were significantly higher than pretest scores (M = 62.37): t(34) = 26.80, p < .001, Cohen’s d = 2.14, N-Gain = 0.60 (moderate gain). The video medium is effective as an initial instructional intervention for Grade 5 fact-and-opinion learning. Future studies should adopt quasi-experimental designs with a control group, incorporate performance-based assessment, and include
Copyrights © 2026