<![CDATA[Short-video platforms now occupy a large share of student leisure time, yet the psychological route linking compulsive viewing to impaired sleep remains poorly specified. This study tested a serial mediation model in which short-video addiction predicts sleep quality through emotion regulation and then academic burnout, treating emotion regulation as a protective resource. A cross-sectional survey collected complete responses from 375 undergraduate students aged seventeen to twenty-five, who reported on twenty-eight Likert items alongside behavioural sleep indicators. Confirmatory factor analysis established the four-factor measurement structure, and the serial mediation model was estimated with 10,000 bootstrap resamples. Short-video addiction predicted poorer sleep quality directly (β = −0.228) and through three indirect routes that were all significant. Academic burnout carried the largest specific indirect effect (β = −0.132), the emotion regulation route carried β = −0.053, and the serial route running through regulation and then burnout carried β = −0.045. Indirect transmission accounted for 50.3 per cent of the total effect of −0.458, and the model explained 39.6 per cent of the variance in sleep quality. Emotion regulation operated as a mediator rather than as a buffer, since its interaction with short-video addiction was not significant. Interventions should therefore target regulatory capacity and academic workload rather than screen time alone.]]>
