Leveraging Digital Devices for Objective Behavioral Health Assessment: Computational Machine Learning Methods for Sleep and Mental Health Evaluation

Massive implementation of digital devices in life is raising concerns over their role in mental well-being. This research considers the correlation between technology usage and mental health by using a sample of 1000 people between 18 to 60 years old. Participants used self-report measures to track their daily time spent on screens, hours spent on social media sites, sleep quality, and numerous other markers of mental well-being. Advanced statistical methods and machine learning models such as Random Forest and Gradient Boosting Machines have been applied to identify and quantify cyberactivity effects. Results The study found some highly significant associations: hours on screen were positively associated with stress levels, (r = 0.67), and low-quality sleep, (r = − 0.42); in contrast, evening use was the worst. Machine learning models had a good predictive accuracy of (up to 85.7%). The most important predictors identified were screen time, hours of sleep, and stress. Clustering techniques unveiled distinct behavioral subgroups where the impact of digital use varies. The results further emphasize the importance of adopting positive digital practices, for example, avoiding exposure in the evening, to mitigate risks to long-term mental health. Although the research draws crucial insights for individuals, policymakers, and designers of information and communication technologies, they do have limitations, where their dependence on self-reports may lead to such bias. Longitudinal studies involving diverse populations remain critical in unraveling and exploring the complex relationship involving the use of technology within the increasingly digital environment.