Esther Mertens

178 | Chapter 7 Summary The current dissertation focused on the effectiveness of universal school-based interventions in improving students’ competencies and preventing problems in the intrapersonal (i.e., feelings, emotions, and attitudes about the self) and interpersonal (i.e., the ability to build and maintain positive relationships with others, to understand social situations, roles and norms, and to respond appropriately) domain. I examined 1) the effectiveness of a specific intervention, Rock and Water (R&W) through a Randomized Controlled Trial, and of universal school-based interventions in general through a meta-analysis, 2) whether intervention dosage and students’ personality traits affected intervention effects, and 3) mechanisms of change and intervention components of universal school-based interventions. Concerning the first aim, universal school-based interventions showed small positive effects in both the intra- and interpersonal domain. Intervention effects found in R&Wwere comparable in magnitude to the effects found for these interventions in general, with strongest intervention effects in the intrapersonal domain. Regarding specific competencies, universal school-based interventions, including R&W, appear unable to improve students’ resilience even though many of these interventions aim to improve this competency. With respect to the second aim, examining intervention dosage showed that R&W was only effective when few teachers were involved. Additionally, students improved most during the first part of the intervention. Hence, the first year might be sufficient to establish change in the assessed competencies and problems. Together these findings imply that for some interventions “less is more”. Regarding students’ personality, the effectiveness of R&Wwas only little affected by students’ personality traits confirming the universal nature of the intervention. However, three moderation patterns did emerge. First, more vulnerable students – based on their levels of certain personality traits – seemed to benefit somewhat more from the intervention than less vulnerable students. Second, extraverted students, but not introverted students, were able to benefit from R&W when many people were involved. Third, personality traits seemed more influential in the intrapersonal domain than in the interpersonal domain. These findings emphasize the importance of examining characteristics of contexts and of participants for determining the extent to which intervention effects can be generalized. Regarding the third aim, analyzing mechanisms of change showed that classmates’ deviant and prosocial modeling and reinforcement did not mediate the effect of R&W on the peer context in the classroom. However, increasing prosocial modeling was related to decreasing victimization, indicating a potential working mechanism. Concerning intervention components, the meta-analysis showed that some components related to stronger effect sizes were rarely implemented (e.g., Individual guidance, Relaxation), whereas some commonly implemented components