48 Chapter 2 Anxiety Fourteen studies (described in 17 papers) included participants with an anxiety disorder, elevated (trait) anxiety symptoms or fear of spiders, of which three studies included participants who additionally had an ASD diagnosis (Wijnhoven et al., 2020), showed elevated levels of externalising problems (Schuurmans et al., 2018) or low mood (McCashin et al., 2022). The studies of Schuurmans and colleagues (Schuurmans, Nijhof, Popma, et al., 2021; Schuurmans, Nijhof, Scholte, et al., 2021; Schuurmans et al., 2020) included participants with clinical levels of post-traumatic symptoms. Participants’ mean age ranged from 9 to 22 years, and the proportion of male participants varied from 11.8% to 83.8%. All studies used a regular RCT design. See Table A.4 (Appendix) for the characteristics and findings of these studies. Given that almost all studies measured anxiety symptoms, this was selected as the primary outcome variable. Effect sizes were calculated on data from thirteen papers. Self-reported anxiety symptoms were used as much as possible, with the exception of McCashin et al. (2022) for which parent-reported anxiety was used. As three papers originated from the same study (Schoneveld et al., 2018), two of them were not included in the forest plot (Schoneveld et al., 2020 reported on secondary outcomes; Wols et al., 2018 focused on effects of one intervention arm). Two other papers were not included in the plot, because anxiety was not measured (Schuurmans, Nijhof, Scholte, et al., 2021) and all intervention arms included applied games (Schuurmans et al., 2020). Eleven papers included two intervention arms, comparing the applied game to an active condition (Beidel et al., 2021; Schoneveld et al., 2018; Schuurmans, Nijhof, Popma, et al., 2021; Schuurmans et al., 2018; Tsui et al., 2021), to a casual game (Haberkamp et al., 2021; Scholten et al., 2016; Schoneveld et al., 2016; Wijnhoven et al., 2020), or to a passive condition (Knox et al., 2011; McCashin et al., 2022). One paper included three intervention arms, namely an applied game, an active condition and a condition including casual games, contributing three effect sizes to the plot (Khanna & Kendall, 2010). In the study of Dennis and O’Toole (2014), four intervention arms were employed. Participants engaged with an applied or placebo version of the game, for either 25 or 45 minutes. For the purpose of this review, we compare the 45-minute gameplay conditions of the applied and placebo version of the game. As shown in Figure A.7 (Appendix), the seven effect sizes comparing an applied game to an active condition ranged from 0.00 to 1.77. Notably, Beidel et al. (2021) found a significant effect favouring the applied game. The five effect sizes comparing an applied game to casual game(s) ranged from 0.11 to 0.46, with only Schoneveld et al. (2016) finding a significant effect in favour of
RkJQdWJsaXNoZXIy MTk4NDMw