Good Games Must Be Good Games: A Mixed-Methods Assessment of User Preferences to Improve Adherence and Engagement with Therapeutic Game Interventions in Adolescents (Preprint)
BACKGROUND Anxiety Disorders (AD) are a major mental health issue in adolescents. While numerous psychological interventions have proven effective in reducing AD, including Cognitive Behavioural Therapy (CBT), and Attention-Bias Modification (ABM), their efficacy is limited by both social and practical barriers (including cost, stigma, and availability), and AD symptomatology (e.g. avoidance-behaviours). As a result, less than 50% of adolescents with AD are estimated to seek treatment, with less than 20% of these ultimately receiving a scientifically validated intervention. Therapeutic games have emerged as an effective medium to promote access to, and engagement with, the therapeutic process. While early results are promising, the mechanisms underlying the ability of therapeutic games to engage this demographic are unclear, with some studies finding initial engagement and reduction in AD symptoms may not be sustainable over time. Despite the importance of player-preferences and participatory design in the development of traditional games for entertainment, little is currently known of adolescent player-preferences in digital health solutions, nor is user-centred design commonplace in this field. Addressing this area has the potential to maximise the impact and economy of gamified interventions, encouraging adherence and engagement over prolonged periods. OBJECTIVE The aim was to conduct an exploratory mixed-methods assessment of adolescents’ player-preferences in therapeutic games for AD. Feedback regarding five digital health games purported to be relaxing or informative about mental health was assessed to establish preferences in game-features and design-styles, and their impact on both player motivation and perceived game-impact. METHODS Sixteen adolescent participants (mean age=17.13, SD=.79) spent at least one hour engaging with at least one of the games and provided feedback via an open- and closed-ended version of a self-report questionnaire. RESULTS Compliance in feedback was high for the quantitative self-report (87.5%) and average for qualitative feedback (56.25%). Across both measures, adolescents generally reported therapeutic games to be an acceptable and engaging medium for delivery of health interventions. A significant preference for games incorporating a variety of tasks was noted (P = .016), and thematic analysis revealed a recurring focus on aesthetics (especially environment colour). Multiple regression analysis demonstrated a range of variables (engagement, functionality, aesthetics, and information) contributed to overall perceptions of game-quality and likeability (F6, 23 = 181.098, P <.001, R2 = .979), but that these variables did not predict perceptions of game impact on the health behaviour (F5, 24 = 4.295, P = .006, R2 = .472). CONCLUSIONS These findings suggest that while adolescents place value on the quality of health information provided in therapeutic games, AD therapeutic game design must also consider design variables, such as game-world aesthetics, game-functionality and game-mechanics, to promote sustained engagement and adherence in this demographic. Implications for further research and participatory therapeutic game design are discussed.