Predicting Mental Health Risks Through Social Media Usage Patterns: A Data-Driven Study of Youth in Odisha

Abstract

The growing popularity of social media among teenagers and young adults has sparked discussions regarding the link between social media and psychological wellbeing and mental illness risk. Social media usage, however, is multi-faceted and spending time online cannot be the only factor that defines psychological outcomes. The proposed research presents a data-based predictive framework for identifying mental health risks among young people from Odisha through social media usage in addition to psychosocial, behavioral and demographic variables. Such variables include frequency of use, time spent online, problem usage, dependence, nighttime use, cyberbullying/victimization, social support, sleep quality, parental control, academic performance, lifestyle and demographics. The specific Odishan data includes a study of 120 undergraduate students at Bhubaneswar whose high social media usage was associated with greater psychological wellbeing and perceived social support compared to low usage subjects. Therefore, high social media usage should not be the sole predictor of mental illness risk. Methods of statistics and machine learning can find the predictors and classify individuals as having mental risks.