From Physical Activity to Online Engagement

Previous studies have demonstrated the advantages of physical behaviour such as physical activity and sleep on mental health and provided an association between virtual behaviour, such as social media use and screen time, and mental health problems. Here, physical behaviour is defined as the data related to the person in the physical real world such as physical activity, sleep, and virtual behaviour is defined as behaviour involving the internet such as social networks, general web browsing, and instant messaging. We believe that a person's physical or virtual behaviour individually may not be the best indicator of their mental health. Current datasets do not include data on both physical and virtual behaviours. Therefore, we seek to run a data collection study that collects both physical and virtual behaviours. Additionally, we investigate if machine learning models that include both physical and virtual behaviour can better predict mental health. This study is conducted by using data collected via a custom-made app. This app is made to run in the background of a user's smartphone collecting physical activity, sleep, location, and audio inferences passively. Additionally, it offers users an ecological momentary assessment (EMA) platform where they may log information about their feelings and other significant occurrences through the Warwick-Edinburgh Mental Wellbeing survey. This will provide us with a ground truth to evaluate our models. We also collect social media data through Instagram and YouTube logs sent by the participants at the end of the study.