Hester van Eeren

Comparative effectiveness of MST and FFT | 5 95 | = often). T-scores were computed and used for analyses. A higher T-score indicates that an adolescent has more problems. Both CBCL and YSR scales were used to measure problem behavior from different perspectives. The Cronbach’s alpha coefficients of the study sample for internalizing, externalizing, and total problem behavior measured with the CBCL were .88, .93, and .96 respectively. For the YSR these coefficients were .92, .90, and .95 respectively. The Cronbach’s alpha coefficients found in the study sample were similar to those reported in the CBCL and YSR manual (i.e., CBCL:.90, .94, and .97, YSR: .90, .90, and .95) (Achenbach & Rescorla, 2001). Finally, until September, 2012, parenting stress was measured with the ‘Nijmeegse Ouderlijke Stress Index’ (NOSI-R) (De Brock, Vermulst, Gerris, Veerman, & Abidin, 2004) in which the primary caregiver completes 42 items on a 4-point scale (ranging from 1 = fully disagree to 4 = fully agree). These items are used to estimate a score for parenting stress wherein a higher score indicates more stress. The reliability coefficient was .95. From October, 2012 onwards, the ‘Opvoedingsbelasting Vragenlijst’ (OBVL) (Vermulst, Kroes, De Meyer, Nguyen, & Veerman, 2012) was used to measure parenting stress. For this measure, the primary caregiver completes 34 items on a 4-point scale (ranging from 1 = not true, to 4 = true). The scores of all items are summed for a total score regarding parenting stress. The alpha coefficient for this measure was .94. Because parenting stress was measured with two different questionnaires, the deviance score of the scales was used to express the level of parenting stress for both questionnaires in one score concerning parenting stress. This was estimated by subtracting the normscore from the score of the adolescent and dividing this by the standard deviation of the norm group. Treatment variables, such as length of treatment and dosage of treatment, were not controlled for in the propensity score since these treatment characteristics are part of the treatment itself and the treatment is adapted to the specific situation of the adolescent and his or her family. Outcome measures Because both FFT and MST are primarily aimed at reducing externalizing problem behavior, this was defined as the primary outcome measure and was measured with the CBCL and with the YSR (Achenbach & Rescorla, 2001). The primary caregiver reported the externalizing problems of the adolescent with the CBCL, while the youth reported this behavior with the YSR. Both measures were completed at the start of and the end of treatment by completing 35 items on a 3-point scale (ranging from 0 = never to 2 = often). T-scores were computed and used for the analyses. A higher T-score indicates that an adolescent has more problems. The alpha reliability coefficient for the current sample at the end of the treatment with the CBCL is .94. For externalizing problems with the YSR, it is .88.