180 Chapter 3 on Sherman, 2015; see Appendix B for written guidelines of all codes) which has been developed to measure the frequency and duration of different ingame play behaviours. The MCS-II includes mutually exclusive codes for the: [1] type of engagement, [2] location in the game-environment, [3] presence of different kinds of fear events, and [4] brightness of the mindlight4 (which could be “none”, “some”, or “total”). In the current study, codes of interest included four engaged in-game play behaviours and two avoidant/safety in-game play behaviours (see Table 1). Engaged behaviours were the frequency of defeat and the duration of exploration, attempts to decloak or attack fear events (“fear attempts”), and the brightness of the mindlight. Avoidant/safety behaviours were the frequency of ceiling light attempts, and the duration of hiding inside a chest. A detailed description of all codes can be found in Appendix B. The first author trained research assistants in the use of Noldus Observer XT and the MindLight Coding System-II. Training took approximately 8 hours and training materials included video data from excluded cases. Coders were blind to the hypotheses. Weekly to bi-weekly recalibration training and reliability checks were conducted to monitor coding and minimize coder drift. Fifteen percent of the total amount of coded video data was independently coded by 2 or more randomly-selected coders. The average reliability was .89 kappa (range .49 – 1.00). Strategy of Analyses The frequency and duration of in-game play behaviours were transformed to frequencies per minute and proportions, respectively, to control for differences in duration of play-sessions. For the brightness of the mindlight, only the durations of the two extremes (“none” and “total”) were used in the analyses. Then, descriptive statistics of the study variables and Pearson correlations between 1) pre-test anxiety scores and in-game play behaviours during the first play-session, and 2) in-game play behaviours during the first play-session and the last play-session were examined. Second, two hierarchical regression analyses (one for engaged and one for avoidant/safety in-game play behaviours) were performed to examine whether changes in in-game play behaviours from the first to the last play-session (i.e., differences in frequencies/proportions calculated as last minus first play-session) were related to changes in anxiety symptoms three months later. The pre-test anxiety score was entered in the first step. The difference in frequency or proportion of the specific in-game play 4 Although the MindWave headset provided a more continuous measure for relaxation through the EEG data, this data could not be matched to the video data. Therefore, the brightness of the mindlight as an indication for relaxation was coded manually.
RkJQdWJsaXNoZXIy MTk4NDMw