Facebook Addiction Disorder Among University Students in Dhaka, Bangladesh: A Cross-Sectional Study Using Bergen’s Facebook Addiction Scale, Neural Network Analysis, and Machine Learning Models

The widespread adoption of social networking sites (SNS) has fundamentally reconfigured social communication, identity expression, and leisure Behaviour in the twenty-first century. Facebook, launched in 2004, has grown to encompass over 2.9 billion monthly active users globally, making it the world’s most utilised social networking platform (Statista, 2023). In Bangladesh, 13.7 million active Facebook users engage primarily via mobile devices, predominantly within the 18–34 age bracket — a demographic profile that overlaps precisely with the university student population (Dhaka Tribune, 2016). This convergence of platform ubiquity, mobile accessibility, and a young, academically engaged user base establishes a fertile context for the emergence of Facebook Addiction Disorder.
Social Networking Addiction (SNA) is conceptualized as the excessive, compulsive use of SNS to the extent that it undermines occupational, academic, social, and psychological functioning. Theoretically, SNA shares mechanistic overlap with substance addictions, engaging dopaminergic reward circuits through variable-ratio reinforcement schedules delivered via social feedback mechanisms: likes, comments, shares, and notifications (Khang, Kim, & Kim, 2013). Bergen’s Facebook Addiction Scale (BFAS), developed and validated by Andreassen, Torsheim, Brunborg, and Pallesen (2012) in a Norwegian sample of 423 university students, operationalizes FAD across six established addiction criteria: salience, mood modification, tolerance, withdrawal, conflict, and relapse. Under the original polythetic scoring convention, individuals scoring ≥4 on at least 4 of 6 items are classified as exhibiting addictive patterns. The present study employs both the original polythetic threshold and an extended rubric (total score 11–30 classified as problematic or addicted) to facilitate comparison with studies that use aggregate scoring approaches.