Page 177 - Demo

Embodied Facial Emotion Processing in Autism and Social Anxiety1756serious head injuries, neurological or other severe internal conditions. Control participants with any first-degree relatives with a history of mental disorders were excluded from the study. An overview for existing comorbidities that did not lead to exclusion can be found in Table S1 in the Supplemental Material. In addition to being matched for gender and age on a group level, we also ensured that the groups did not significantly differ in their level of education, indexed by the years of education (AS group: M = 17.68, SD = 5.24; SA group: M = 16.48, SD = 2.92; NC group: M = 16.7, SD = 3.29) and intelligence indexed by the Multiple Choice Vocabulary Intelligence Test (MWT-B; Lehrl, 2005; see Supplemental Materials for additional information). Group differences in diagnostic and self-reported trait variables are outlined in Table S3 in the Supplemental Material.Clinical participants were recruited from different in- and outpatient units in the LVR-University Hospital in Essen. Moreover, nearby therapy centres as well as patient groups (e.g. support groups, online forums) were contacted. Newspaper and online advertisements (e.g., of the Scientific Society Autism Spectrum WGAS) were published. The sample size was determined based on a power analysis using the simr package in R (P. Green & MacLeod, 2016). More specifically, we used an observation from our previous study in a student sample as basis, namely lower confidence in general emotion recognition with higher social anxiety trait levels (%u03b2 = - 0.13, Folz et al., 2023). A sample size of 40 participants was deemed sufficient to detect a medium effect (-0.5) with 100% power and a small effect (-0.26) with 77% power. Importantly, this power analysis covered one specific effect and we only managed to recruit 27 individuals with social anxiety during the study period (September 2021 %u2013 October 2023, including the COVID-19 pandemic). Therefore, we decided to additionally employ Bayesian analyses to identify whether our sample was large enough to find convincing evidence for the alternative or null hypothesis in our group comparisons of interest.Written consent was obtained from all participants prior to participation. The current study was part of a larger research project including additional tasks and measurements (see Procedure section), which encompassed two testing days of 2-2.5 hours duration each. After completion of all tasks on the second testing day, participants were given verbal and written debriefing about the study and received a fixed reimbursement of 50%u20ac for their travel costs and varying bonus payments depending on outcomes in another task (0 %u2013 2.5%u20ac). The design of