Esther Mertens

58 | Chapter 3 Control condition (slope = -.41). In both the Light and Standard condition the strongest improvements were again in the first year. No intervention effects were found for resilience and for students in the Plus condition (see the supplementary material for the slopes of all conditions and outcomes in the intrapersonal domain). When comparing students in the Light condition with students in the Standard condition on change over time in internalizing behavior, the unconstrained model fitted the data significantly better, Δχ 2 SB (1) = 6.60, p = .010, than the constrained model (i.e., slopes constrained to be equal across the conditions). Students in the Light condition showed a stronger decrease in internalizing behavior than students in the Standard condition (slope Light  = -.60; slope Standard  = -.53). Overall, students in the Light condition seemed to have a slightly more beneficial change over time in the intrapersonal domain with the intervention’s primary impact early in the intervention. Interpersonal domain. Students in the Light condition showed a more beneficial trajectory of change for aggression compared to students in the Control condition (see Figure 2). The effect size was small (Cohen, 1988). Students in the Light condition remained relatively stable over time (slope = .09), whereas students in the Control condition showed an increase in aggression (slope = .56). The strongest intervention effect was found in the first year of the intervention. Furthermore, two trends suggested that students in the Light condition also had a slightly more beneficial trajectory of change for interpersonal relations in the class and bullying compared to students in the Control condition (see Figure 2). Students in the Light condition showed a less steep decrease in positive interpersonal relations in the class (slope = -1.64) and a less steep increase in bullying (slope = 2.34) than students in the Control condition (slope Interpersonal relations  = -2.03; slope Bullying  = 2.80). Both effect sizes were small (Cohen, 1988) with the strongest effects found in the first year. No intervention effects were found for externalizing behavior and victimization. Additionally, no intervention effects were found for students in the Standard and Plus conditions (see the supplementary material for the slopes of all conditions and outcomes in the interpersonal domain). Overall, the Light condition also seemed to have a slightly more beneficial change over time in the interpersonal domain with, again, the strongest intervention effect in the first year of the intervention.