Esther Mertens

154 | Chapter 7 focusing on these questions future research can further deepen our understanding of universal school-based interventions and possibly increase their effectiveness. Students who are vulnerable to develop problems are an important subgroup to study, especially in universal interventions. Universal interventions specifically aim to prevent the development of problems (Nehmy & Wade, 2014). As vulnerable students have an increased risk of developing problems, it is eminent that particularly those students are able to benefit from intervention. The present study found stronger intervention effects for students identified as vulnerable based on their levels of personality traits. However, it remains unclear to what extent these advantageous effects can be generalized to other subgroups of vulnerable students (Rowe & Trickett, 2018), such as students with elevated levels of problems at baseline or victimized students (e.g., Kaufman, Kretschmer, Huitsing, & Veenstra, 2018; Multisite Violence Prevention Project, 2008). In order to determine whether universal interventions are indeed able to particularly target the more vulnerable students, it is pivotal to examine intervention effectiveness for different subgroups of vulnerable students. Furthermore, unraveling mechanisms of change is eminent in order to link intervention techniques with students’ development (Michie et al., 2009), even when the intervention effects are small. In the present study, however, the mediators that were tested (i.e., modeling and reinforcement) did not appear to be mechanisms of change within R&W. Inmy future research I tend to examine self-control, self-reflection, self-esteem, and emotional self-regulation as mechanisms of change. The theory of the R&W house (Ykema 2002; 2018) states that R&W improves these competencies in students which, in turn, would enable these students to develop themselves in the intra- and interpersonal domains, suggesting mediation. The results of the present study imply that R&W is indeed able to improve students’ self-control and emotional self-regulation, and showed promising (but not significant) changes in self-esteem (but not in students’ self-reflection). Furthermore, these competencies have been related to competencies and problems in the intra- and interpersonal domain by previous research (e.g., Hughes et al., 2020; Mann, Hosman, Schaalma, & De Vries, 2004; Otten, Barker, Maughan, Arsenaeault, & Engels, 2010), making also the relation between the mediators and outcomes plausible. Hence, analyzing these potential mechanisms of change could contribute to understanding how interventions in general can foster students’ development in both the intra- and interpersonal domain and, at the same time, examines the theory of the R&W house. Although universal school-based interventions show positive effects on students’ competencies and problems in the intra- and interpersonal domain, these effects are small, indicating the necessity to improve the effectiveness of these interventions. The results of my meta-analysis suggests that there is potential to improve these interventions as some commonly used components might be ineffective and other potentially effective components are rarely implemented. However, the relations between components and intervention effects in the meta-analysis are based on correlations. Future research needs to study these components explicitly to determine