Esther Mertens

| 149 General Discussion sufficient to establish the effects in the assessed competencies and problems. Previous research also found that intervention effects decrease over time (e.g., Cuijpers, 2000; De Mooij, Fekkes, Scholte, & Overbeek, 2019) and illustrated that participants who benefit from an intervention often show these improvements early in the intervention (e.g., Lutz et al., 2014; Tadić et al., 2010). This seemingly typical decline in intervention effects may be related to participants’ expectations and motivation since these are moderators of intervention effects (Ebert et al., 2013: Philips & Wennberg, 2014). Although not reported in one of the chapters in this dissertation, I assessed students’ attitudes towards R&W and analyzed changes in their expectations and motivation to participate. The results showed that students’ expectations and motivation to participate in the intervention indeed declined, which could have contributed to the decreases in intervention effects. Perhaps a 2-year intervention is simply too long to maintain the interest of students. Together, the findings regarding intervention dosage support previous research stating that more extensive and longer interventions are not necessarily more effective, but that sometimes “less is more” (e.g., Cuijpers, 2000; Durlak, Weissberg, Dymnicki, Taylor, & Schellinger, 2011). Concerning characteristics of the student population, students’ personality traits had generally little effect on intervention effects, confirming the universal nature of the intervention. However, three patterns of moderation did emerge. First, in line with the Risk moderation hypothesis (Spoth, Shin, Guyll, Redmon, & Azevedo, 2006), a compensatory effect was found for more vulnerable students. More specifically, students who were vulnerable to develop problems in the intra- or interpersonal domain – based on their levels of certain personality traits – benefitted most from the intervention. For instance, without intervention, students with high levels of Neuroticism improved less in psychological wellbeing than students with lower levels of this trait. However, when these vulnerable students received the intervention, they improvedmore in psychological wellbeing than less vulnerable students (i.e., students with low levels of Neuroticism). This pattern suggests that R&W might be able to prevent the development of problems indicating a true prevention effect (Nehmy & Wade, 2014). The second pattern suggested that high levels of Extraversion enabled students to benefit from an intervention in which many people are involved in the implementation. Extraverted students benefitted from R&W regardless of the number of involved people, whereas the involvement of more people (i.e., Standard and Plus conditions) seemed to counteract the intervention effects for less extraverted students. Perhaps the learning processes of extraverted students, characterized by cooperation and group work rather than internal processing of information (Felder, Felder, & Dietz, 2002), fits well with an intervention approach in which interactions with others are important, enabling them to benefit from the intervention. The third pattern indicated that personality traits affected intervention effects in the intrapersonal domain more than in the interpersonal domain. Personality traits can influence how one wants to feel and why (Hughes, Kratsiotis, Niven, & Holman, 2020) and might therefore be more related to one’s feelings, emotions, and attitudes (i.e., the intrapersonal domain), 7