Esther Mertens

148 | Chapter 7 reported in this dissertation) suggest that self-reflection might be a different type of competency since self-control, self-esteem, and emotional self-regulation correlated strongly together ( r = .50 - .55), but weakly with self-reflection ( r = -.15 - -.25). Perhaps self-reflection is cognitively too demanding for most prevocational students as it requires abstract thinking on the metacognition level (Sauter, Heyne, Bloöte, Van Widenfelt, & Westenberg, 2010). Alternatively, self-reflection might be stimulated through a different intervention method than self-control, self-esteem, and emotional self-regulation. Therefore, the next step is to examine how students’ self-reflection can be stimulated and whether these four competencies might function as mediators given that the theory of the “Rock &Water house” suggests these competencies to be working mechanisms of the intervention (Ykema 2002; 2018). This puts the theory of the intervention to the test and increases understanding of the process of change which might be helpful for optimizing interventions (Kazdin, 2007). Overall, the effectiveness of R&W is comparable with other universal school-based interventions addressing students’ development in the intra- and interpersonal domains. It is difficult to determine the practical relevance of the improvements seen in the students. Even though the intervention effects are small, which is to be expected in universal interventions as these target a mainly healthy population (Nehmy &Wade, 2014), the potential impact can be large. If many individuals change their behavior just a bit, this could lead to large benefits in society (Greenberg & Abenavoli, 2017). Yet, the small intervention effects indicate that there is room to optimize universal school-based interventions in general and, more specifically, R&W. Aim 2: Heterogeneity in Contexts and in Population The second aim of this dissertation was to examine whether characteristics of the context (i.e., intervention dosage: Ecological focus and time span) and student population (i.e., personality traits) affected intervention effectiveness. Regarding the first aspect of intervention dosage, the findings showed that an intervention’s ecological focus could affect intervention effects. R&W was only effective when the intervention had a narrow ecological focus, that is, when it was solely implemented during the intervention lessons. Hence, the circumstances under which an intervention is implemented can have a large influence on the intervention’s effectiveness. More importantly, neglecting the context of an intervention could lead to drawing erroneous conclusions about the intervention’s effectiveness. For instance, if I only examined R&W under “standard” conditions, I may have concluded that the intervention might be hardly worthwhile to implement. The second aspect of intervention dosage, that is, the time span of intervention, also affected intervention effects. Students improved most during the first part of the intervention, whereas improvements in the second year were neglectable. This finding suggests that implementing only the first year of the intervention might be