Social Media Use and Student Disengagement: A Systematic Review of Distraction and Technostress

As algorithmically curated social and mobile platforms become embedded in students’ everyday learning environments, concerns about AI-driven distraction, persuasive engagement design, and technostress have become increasingly central to higher-education research and practice. While prior work has examined problematic social media use and digital wellbeing, a focused, PRISMA-guided synthesis that connects specific platform design mechanisms (e.g., notifications, personalization, infinite scroll, social rewards) to students’ cognitive, emotional, and academic outcomes remains fragmented across disciplines and measures. This systematic literature review addresses that gap by following PRISMA 2020 guidance and synthesizing evidence from 34 included studies selected through a transparent screening process (records identified: EBSCO n=1,374; Web of Science n=1,125; Scopus n=846; total n=3,345; duplicates removed n=1,072; records screened n=2,273; excluded at title/abstract n=945; full-text assessed n=1,328; excluded at full-text n=1,294). Across the included literature, we observe recurring patterns linking engagement-oriented design features and algorithmic curation with attention fragmentation, FoMO/compulsive checking, and technostress-related strain, alongside downstream associations with academic performance, wellbeing, and self-regulation. At the same time, the evidence base is constrained by conceptual inconsistency (overlapping constructs and definitions), measurement heterogeneity, and an overreliance on cross-sectional self-report designs, limiting causal inference and comparability across studies. To support reproducibility and cumulative science, we provide an open screening dataset and a study-level codebook (with copyright-sensitive abstract text removed from public files) and outline a mechanism-to-outcome synthesis structure to guide future research toward clearer operationalization of platform mechanisms, stronger designs (longitudinal/experimental), and more standardized reporting and evaluation.