Using data mining to discover new patterns of social media and smartphone use and emotional states

Abstract

Social media and smartphone use are strongly linked to users' emotional states. While numerous studies have established that fear of missing out (FOMO), boredom, and loneliness predict social media and smartphone use, numerous other studies have concluded that social media and smartphone use negatively impact these emotional states (i.e., FOMO, boredom, and loneliness). Phubbing (phone snubbing), which is the act of ignoring a physically present person in favour of a smartphone, is associated with both social media and smartphone use and users’ emotional states. Much of the above research, however, has adopted the traditional hypothesis testing method. So far, limited work has been done using data-driven approaches. This paper uses data mining techniques to uncover previously unknown patterns about social media and smartphone use, phubbing, and users' emotional states based on two existing datasets originating from online questionnaires facilitated through social media. Novel patterns related to FOMO, loneliness, boredom, and phubbing are discovered and explored in detail. The study also demonstrates the usefulness of the data-driven approach and establishes it as a valid alternative to the hypothesis-driven approach to investigating social media and smartphone use, phubbing, and users' emotional states.