Aniek Wols

242 Chapter 5 Trailer Manipulation Check To examine whether the two trailers induced different expectations, participants answered two open questions: “What were your impressions of MindLight right after you watched the trailer?”, and “What did you expect from MindLight based on the trailer (before you played the game)?”. In addition, participants answered the following Yes/No question: “Did you notice that the message in the trailer were focused on [game enjoyment (ENT-condition)] or [on how MindLight could help people who feel stressed / anxious or have some mental health difficulties (MH-condition)]?”. Statistical Analyses First, trailer manipulation and randomisation were checked. Bootstrapped Pearson correlations between study variables are reported in Table 2. Bootstrapped paired t-tests were used to examine changes in affect, stateanxiety and arousal. The remaining preregistered research questions were examined within a (hierarchical) regression framework (controlled for high correlations; see preregistration) (IBM Corp, 2017). Univariate outliers were winsorized to ± 3.5 standard deviation from the mean before conducting the analyses. Because some study variables were not normally distributed, all regression models were bootstrapped with n = 1000 samples. For GSR, univariate outliers were winsorized to ± 3.5 standard deviation from the mean, both for the overall GSR percent change value and the timebin values. The exploratory analyses for GSR included 1) a bootstrapped onesample t-test to examine whether overall GSR percent change during gameplay was higher than zero, 2) bootstrapped regression analyses to examine differences between the experimental conditions and the interactions with motivation to change and anxiety symptoms on overall GSR percent change, and 3) a Repeated Measures ANOVA with the six 10-minute timebins to examine (polynomial) changes in GSR during gameplay, with experimental condition as between-subjects factor.

RkJQdWJsaXNoZXIy MTk4NDMw