<![CDATA[The limitations and conditions of teachers who have difficulty accessing or developing cognitive stimulus learning media for abstract and complex material are a challenge in themselves in implementing the discovery model. This study aims to evaluate an AI-based stimulus website developed to support the discovery learning model. learning in science learning through user perception, motivation, metacognition, and student engagement. This study uses a quantitative approach using Partial Least Squares Structural Equation Modeling (PLS-SEM). The results of the analysis show that the influence of learning Comprehension on Engagement has a coefficient of β = 0.437 (p = 0.002) and on Motivation Metacognition, β = 0.446 (p = 0.000), indicating that learning comprehension significantly increases student engagement and motivation. Perception of the AI website also has a strong influence on engagement (β = 0.550, p = 0.001) and motivation metacognition (β = 0.477, p = 0.002). While video quality does not have a significant effect on engagement (β = -0.052, p = 0.607) or motivation metacognition (β = 0.034, p = 0.793), indicating that psychological and cognitive aspects are more important than technical aspects of video in AI-based learning. Positive perception and conceptual understanding through AI-based websites significantly increase students' motivation, metacognition, and engagement, while the technical quality of the media does not play a dominant role. These results provide recommendations for developers and educators to prioritize interactive designs that support metacognitive reflection and active engagement of students as independent learners in discovery learning that uses AI-based stimuli.]]>
Copyrights © 2026
