188 Chapter 3 three-month follow-up assessment (when children had not played the game for three months). These findings show that what children learned and practiced in the game led to actual changes in their real-life (self-reported) behaviours. Most CBT approaches are not able to close this gap (Granic et al., 2014; Kendall, 2011) and only a few game-based interventions have examined these effects (e.g., Girard et al., 2013). Moreover, by including a long-term follow-up, the present study is one of the few longitudinal studies in the field of applied video games (Girard et al., 2013; Granic et al., 2014; Primack et al., 2012). Limitation and Suggestions for Future Research There are a few limitations that should be mentioned when interpreting the findings and future research should attempt to overcome. The first limitation is that no in-game play behaviours representing the third technique (i.e., attention bias modification), that was initially incorporated in MindLight, were included in the current study. Because of how the game was designed, these behaviours were too dependent on the specific puzzle and the current location of the player and these locations varied widely across players, leaving comparisons across children unreliable. For example, a puzzle success on an easier puzzle would be different from a puzzle success on a more difficult puzzle, which requires more puzzle attempts and also the ability to distinguish different varieties of positive faces and negative faces. However, evidence for the effectiveness of attention bias modification in the literature is mixed. Some research has shown that attention bias modification is capable of reducing anxiety (Bar-Haim, 2010; Bar-Haim et al., 2011; Hakamata et al., 2010), whereas other recent studies – that were published after MindLight was developed - failed to show this (for an overview see Clarke et al., 2014; Cristea et al., 2015). Clarke and colleagues (2014) argue that the studies finding no effects of attention bias modification failed to manipulate attention and therefore found no effects on the outcome measures. Future research should first test whether the puzzles in MindLight actually are able to modify attentional bias before in-game puzzle-solving behaviours are examined in relation to changes in anxiety. A second limitation is that despite the fact that inactivity has good face validity for representing avoidant/safety behaviour because it reduces exposure to fear events and literally limits engagement with the game mechanics in MindLight, the inactivity code in the MindLight Coding System-II is not able to distinguish between actively avoiding the game or being inactive because the child is waiting for a research assistant to answer a question. Relatedly, the technical problem code might have been confounded too, because this code was used – but not exclusively – when the game was paused. It is possible
RkJQdWJsaXNoZXIy MTk4NDMw