Esther Mertens

| 75 Role Personality in School-Based Intervention Figure 1 . Flow chart. Assessed for eligibility ( n = 37 schools) Excluded ( n = 24 schools) • Did not meet inclusion criteria ( n = 6) • Declined to participate ( n = 10) • Other reason ( n = 8) Randomized ( n = 13 schools) Allocated to Light condition ( n = 4) Allocated to Standard condition ( n = 3) Allocated to Plus condition ( n = 3) Allocated to Control condition ( n = 3) 1 School dropped out, replaced by other school T1 Baseline • Included students ( n = 373) • No consent student ( n = 9) • No consent parent ( n = 3) T1 Baseline • Included students ( n = 303) • No consent student ( n = 8) • No consent parent ( n = 4) T1 Baseline • Included students ( n = 249) • No consent student ( n = 0) • No consent parent ( n = 1) T1 Baseline • Included students ( n = 374) • No consent student ( n = 3) • No consent parent ( n = 7) 1 st year • 1 trainer dropped out in 1 st year ( n = 45 students) • Left school n = 40 2 nd year • 1 class not received R&W ( n = 18, 16 completed all questionnaires, 2 left school) • Left school n = 15 1 st year • Left school n = 28 2 nd year • 1 class not received R&W ( n = 10, 8 completed all questionnaires, 2 left school) • Left school n = 23 1 st year • Left school n = 17 2 nd year • Left school n = 6 1 st year • Left school n = 31 2 nd year • Left school n = 18 T4 post measurement • Completers n = 271 (73%) T4 post measurement • Completers n = 250 (83%) T4 post measurement • Completers n = 226 (91%) T4 post measurement • Completers n = 325 (87%) Adolescents completed digital questionnaires at four time points: At baseline (T1; October/November 2017), after the first year of the intervention (T2; March/April 2018), before the start of the second year of the intervention (T3; October 2018), and at post measurement (T4; January 2019). Questionnaires were conducted by trained research assistants. Parents completed a digital questionnaire at baseline. 4