“Quiet Screens, Loud Minds”: Classroom-Based Assessment of Nocturnal Social Media Use, Sleep Quality, and Anxiety Among Middle Adolescents in Ghana

Abstract The study examined the relationship between nocturnal social media use, sleep quality, and anxiety levels among middle adolescents in Ghana, with a focus on whether sex moderates these associations. A sample of 1,200 adolescents from southern Ghana participated, with data collected through self-report surveys and teacher assessments. Pearson’s correlation analysis revealed a significant negative correlation between nocturnal social media use and sleep quality (r = − 0.45, p < 0.01), indicating that increased screen time before sleep was associated with poorer sleep quality. Poor sleep quality was also positively correlated with higher anxiety levels (r = 0.38, p < 0.01). Multiple regression analysis showed that classroom functioning, as assessed by teachers (attention, participation, mood), significantly predicted anxiety levels (β = 0.22, p < 0.05) and sleep quality (β = − 0.30, p < 0.01), with sleep disturbances mediating the relationship between nocturnal social media use and anxiety (path coefficient = 0.23, p < 0.01). Importantly, moderated path analysis revealed that sex significantly influenced these relationships, with stronger negative effects observed among males. Structural path analysis revealed that nocturnal social media use significantly predicted sleep disruption (β = 0.56, CR = 11.20, p < .001), and sleep disruption strongly predicted anxiety symptoms (β = 0.61, CR = 10.17, p < .001), confirming a robust mediating pathway. Even when accounting for sleep disruption, nocturnal social media use still had a direct effect on anxiety (β = 0.34, CR = 4.86, p < .001), indicating partial mediation. The indirect effect of sleep disruption on anxiety was also significant (β = 0.34, CR = 8.50, p < .001). Teacher-rated classroom functioning contributed notably: higher attention (β = − 0.42, CR = − 7.00, p < .001) and participation (β = − 0.31, CR = − 5.17, p < .001) were associated with less sleep disruption, while negative mood predicted higher anxiety symptoms (β = 0.37, CR = 6.17, p < .001). Sex significantly moderated the pathways. Male adolescents were more prone to sleep disruption (β = 0.29, CR = 4.14, p < .001), and the interaction between sex and nocturnal social media use was significant (β = 0.18, CR = 3.60, p = .001), suggesting that males are more vulnerable to the sleep-disrupting effects of social media. The study recommended that stakeholders should collaborate to implement sleep hygiene programs and limit nighttime screen use, with attention to gender-sensitive interventions that promote emotional well-being and academic success.