VIDEO GAME PLAY, SOCIAL INTERACTIONS AND FRIENDSHIP QUALITY IN ADOLESCENTS: A LATENT CLASS ANALYSIS

Rationale. Video games and the Internet are an indispensable part of teens’ lives and are tightly woven into their daily routines. The ever-changing nature of games and online settings poses challenges for public health researchers who categorize exposures to determine how they affect health. Previous research has identified a range of correlates and outcomes of problematic gaming; many of these are shared with addiction to Internet activities such as social networking and instant messaging, which are also very popular with adolescents (D. L. King, Haagsma, Delfabbro, Gradisar, & Griffiths, 2013; D. J. Kuss, Griffiths, Karila, & Billieux, 2013). Since most youth use a variety of Internet social applications at once (Rideout, Foehr, & Roberts, 2010), it is important to consider how these concurrent activities affect the nature, course, and outcomes of problematic gaming. Given the overlapping nature of Internet use, some forms of gaming, and social interactions, the outcomes of heavy or excessive gaming are likely vary as a function of online social interactions (e.g., those occurring through instant messaging, social media use and chat; S. Caplan, Williams, & Yee, 2009; Ng & Wiemer-Hastings, 2005). This thesis adds to the literature by first providing a systematic examination of factors related specifically to problematic gaming (as opposed to problematic Internet use) in representative samples of adolescents, then conducting two empirical studies to determine (1) how patterns of self-reported problematic gaming symptoms, video game and social Internet use help distinguish adolescents whose high levels of gaming are associated with high problematic gaming symptoms and (2) how these patterns are associated with psychosocial well-being, and how these associations change when online and offline friendship quality is considered. Methods. We used survey results from 9738 adolescents in the 2009-2012 waves of the Monitor Internet and Youth study (Meerkerk, van den Eijnden, & van Rooij, 2006), a yearly survey on technology use and addiction in the Netherlands. We conducted a latent class analysis using indicators of problematic gaming symptoms (as measured by the Videogame Addiction Test; van Rooij, Schoenmakers, van den Eijnden, Vermulst, & van de Mheen, 2012a) and high levels of use (4+ hours per day for 6 -7 days per week) of social networking, instant messaging and three types of video games. This analysis was then extended to a multivariate latent class regression that examined associations between the estimated latent classes and depression, social anxiety, loneliness and low self-esteem. Finally, we evaluated moderation by friendship quality. Results: In both boys and girls, we identified several classes of gamers with varying levels of problematic gameplay. Classes were generally split into those having high probabilities of highly engaged online social interaction and those having lower probabilities, and classes with more online social interaction showed fewer symptoms of problematic gaming at the same levels of gaming. Classes with higher self-reported problematic gaming were associated with lower values of self-esteem and more social anxiety overall, whereas boys and girls estimated to be in the Social Engaged Gamers classes had less loneliness and social anxiety. All classes with high levels of gaming also reported more depressive symptoms. However, these associations changed for some subgroups when interactions with online and offline friendship quality were considered. Specifically, for girls who had good online friendships but poor real-life friendships, the chance of membership in the Social Engaged Gamers class compared to the Average class decreased as depression scores increased. High self-esteem was negatively associated for most boys who were At-Risk Social Gamers, but for boys who had high quality online and offline friendships, the relationship was positive. In addition, social anxiety was only associated with the most symptomatic Problematic Gamers class for boys with high-quality online friendships, but was associated with both heavy gaming classes for girls who had high-quality online and real-life friendships. Conclusion: Adolescent gamers who were more socially active online reported lower distress related to gaming. Although all highly engaged gamers reported more depression than average classes, high-quality online friendships negated this finding for girls who reported low-quality real-life friendships. Exploring moderation also revealed that associations between social anxiety and certain classes were much stronger for those with high-quality online friendships, and self-esteem was higher for some social gamers if both online and offline friendship quality were high. Online social interaction, especially when it is associated with high-quality online friendships, may have positive implications for some heavy gamers but may be associated with more negative psychological functioning in others