Esther Mertens

| 79 Role Personality in School-Based Intervention Interpersonal domain Interpersonal relations in the class. Negative social exchanges between classmates, the extent to which adolescents feel comfortable around their classmates, and the unity and inclusiveness among classmates was measured using the Classroom Peer Context Questionnaire (Boor-Klip, Segers, Hendrickx, & Cillessen, 2016). The questionnaire contains 12 items (e.g., “In this class students like each other.”) rated on a 5-point Likert-type scale (1 =  totally not true to 5 =  completely true ). Some items were recoded with high scores indicating more positive interpersonal relations in the class (Cronbach’s α = .80 - .85). Externalizing behavior. The presence of externalizing problems was assessed with the externalizing subscale of the short version of the YSR (Chorpita et al., 2010). The subscale contains 6 items (e.g., “I destroy things.”) rated on a 3-point Likert-type scale (0 = never to 2 = often ; Cronbach’s α = .65 - .79). Aggression. Aggression was measured with the Reactive and Proactive Aggression Questionnaire (Dodge & Coie, 1987). The questionnaire contains 6 items (e.g., “If they tease me, I get angry.”) answered on a 5-point Likert-type scale (1 = never to 5 = almost always ; Cronbach’s α = .65 - .83). Bullying and victimization. The frequency of bullying and experienced victimization was measured with the 2 global items of the Olweus Bully/Victim Questionnaire (Solberg, & Olweus, 2003): “How often have you taken part in bullying others?” and “How often have you been bullied?” answered on a 5-point Likert-type scale (1 = never to 5 = almost always ). The items were preceded by a definition of bullying. Moderator Personality. To assess adolescents’ personality, parents completed the Quick Big Five (Goldberg, 1992). This questionnaire contains 30 items (e.g., nice, sympathetic, organized) representing the 5 personality traits: Extraversion, Agreeableness, Conscientiousness, Neuroticism, and Openness to experiences, with each trait assessed by 6 items. Parents indicated to which extent that characteristic suites the adolescent on a 7-point Likert-type scale (1 = not at all to 7 = very well ; Cronbach’s α = .80 - .91). Statistical Analyses Analyses were conducted in M plus version 8.2 (Muthén & Muthén, 2010) with an intention-to-treat approach, i.e., all adolescents who participated at baseline were analyzed regardless whether they actually received the intervention or not. The nested structure of the data was taken into account by using the complex sample cluster 4