The pervasive adoption of algorithmic short-form video architectures—predominantly TikTok, YouTube Shorts, and Instagram Reels—has fundamentally transformed the information consumption habits of secondary school students. Despite growing concerns regarding cognitive fragmentation, empirical evidence quantifying the direct and indirect impacts of short-form video consumption on autonomous academic behaviors remains scarce. This empirical study examines the structural relationships between short-form video exposure, sustained attention span, and self-directed learning (SDL) readiness among high school students. Utilizing a cross-sectional correlational design, data were collected from N=480 students (grades 10–12, ages 15–18) across six secondary schools. Validated psychometric instruments were administered, including the Short-Form Video Addiction Scale (SFVAS), the Mindful Attention Awareness Scale for Adolescents (MAAS-A), and the Self-Directed Learning Readiness Scale (SDLRS). Hierarchical regression and bootstrap-based mediation analyses (5,000 resamples) revealed that daily short-form video consumption significantly and negatively predicted sustained attention (β=-0.462,p<.001) and self-directed learning readiness (β=-0.388,p<.001). Furthermore, sustained attention exerted a robust positive effect on self-directed learning (β=0.441,p<.001) and partially mediated the relationship between video consumption and learning readiness (indirect effect β=-0.204,95% CI -0.258,-0.154), accounting for 52.5% of the total effect. These findings provide empirical confirmation of cognitive load displacement and executive function attenuation in hyper-stimulating digital environments, offering concrete implications for pedagogical design, cognitive scaffolding, and secondary school digital literacy frameworks.
Keywords: Short-form video, Attention span, Self-directed learning (SDL), Cognitive load theory, Adolescent neurodevelopment, ICCMETS 2026.