Esther Mertens

86 | Chapter 4 Neuroticism. Neuroticism was negatively related to psychological wellbeing, regardless of condition, indicating that higher levels of Neuroticism predicted a less steep increase in psychological wellbeing over time. In addition, Neuroticism moderated intervention effects on psychological wellbeing in the Light condition and on resilience in the Light and Standard conditions. Adolescents with high levels of Neuroticism improved more on psychological wellbeing ( d ­ low  = -.33, d average = -.07, d high  = .74). Contrarily, the intervention was less successful than CAU in improving resilience for adolescents with high levels of Neuroticism (Light: d low  = .06, d average  = -.02, d high  = -.89; Standard: d low  = .52, d average  = .03, d high  = -.60). No moderation effects were found for the other outcomes or in the Plus condition. Openness to experiences. Openness to experiences was positively related to resilience and negatively related to sexual autonomy, regardless of condition. This means that more Openness to experiences was related to a less steep decrease in resilience and a less steep increase in sexual autonomy over time. Furthermore, Openness moderated intervention effects on internalizing behavior in the Light condition. Adolescents with average levels of Openness decreased more on internalizing behavior ( d low  = -.12, d average = .41, d high  = .06). No moderation effects were found for any of the other outcomes or in the other conditions. Discussion The present study expanded previous research by examining the effects of personality traits on trajectories of change during a universal intervention across a broad range of competencies and problems in the intra- and interpersonal domains. Overall, the intervention appeared moderately effective in improving several aspects of both adolescents’ intra- (i.e., psychological wellbeing, sexual autonomy, and internalizing behavior) and interpersonal (i.e., externalizing behavior and aggression) domains. Moderate intervention effects were expected based on meta-analyses of school- based interventions (e.g., Cohen’s d = .22 - .27, Durlak, Weissberg, Dymnicki, Taylor, & Schellinger; Cohen’s d = .10 - .25, Mertens, Deković, Leijten, Van Londen, & Reitz, 2020). Personality traits affected only a few intervention effects which confirms the universality of the intervention. In general, three patterns of moderation emerged. First, therewere indications that adolescents vulnerable to develop problems in the intra- and interpersonal domains – based on their levels of certain personality traits – benefitted more from the intervention than less vulnerable adolescents. Second, Extraversion appeared to be an importantmoderator of intervention effects in interventions requiring sociability from participants. Third, personality traits seemed to affect competencies and problems in the intrapersonal domain somewhat more than in the interpersonal domain, as predictors as well as moderators of intervention effects.