3 189 IN-GAME PLAY BEHAVIOURS PREDICT IMPROVEMENTS IN ANXIETY that some children paused the game to actively avoid fear events in the game. Future research should try to separate these different behaviours, such that codes become as clear as possible and might show relations to changes in anxiety symptoms. Finally, the current study did not test mediators or moderators. It is possible that associations between (multiple) in-game play behaviours and changes in anxiety symptoms are mediated or moderated by other factors, such as game expectancies, game experience, motivation, and enjoyment, which are known to be important predictors of treatment outcome (Castonguay et al., 1996; Ferguson & Olson, 2013; Przybylski et al., 2010). Future research might want to examine these effects in a larger sample to get insight into why and for whom MindLight works best. Practical Implications Notwithstanding the aforementioned limitations, the current study has three important implications. The first implication applies to clinical research on anxiety. Results showed that in-game behaviours representing the exposure principle predicted improvements in anxiety. These findings underscore the importance of this technique in anxiety disorders and show that applied games can be used to reduce anxiety symptoms. Using game-based interventions contribute to the movement of tailoring and personalizing treatments (Chorpita & Weisz, 2009) by engaging and motivating children that might not like the rather didactic-based approach of traditional CBT-treatments (Gosch et al., 2006) and by providing rich practice opportunities for children that find it difficult to use newly-acquired knowledge in real-life situations (Granic et al., 2014). The second implication relates to designs aimed at improving both gameplay experiences and intervention outcomes. Because results demonstrated that the MindLight Coding System-II is useful for distinguishing between engaged and avoidant/safety in-game play behaviours, the coding system could be integrated into the game itself. In-game play behaviours can then be tracked automatically and be used as a measure for determining children’s progress. Measurements of in-game behaviours can be used to dynamically adjust the game to the player’s actions, diverse needs and learning paces (Bakkes et al., 2012; Bakkes et al., 2014). For example, the game may provide more hiding spaces in the beginning of the game for more anxious children (cf. Milosevic & Radomsky, 2008; cf. Rachman et al., 2008), or provide more exposures to fear events when children are able to stay relaxed and calm in the face of less/easier fear events. Additionally, such measurements
RkJQdWJsaXNoZXIy MTk4NDMw