Page 85 - Age of onset of disruptive behavior of residentially treated adolescents -Sjoukje de Boer

Kappa: .79) has been determined between ratings independently done by two researchers on a random subset of the sample (n=11), based on the available multi- informant (adolescent, parent, and therapist), and multi-method (self-report, interview) information.
Dropout
Dropout was determined in line with the preferred definition of dropout of De
Haan et al. (2013), based on how treatment was terminated according to the therapist 5 involved. This resulted in three groups: those who terminated treatment in accordance
with the therapist (i.e., completers), those who terminated treatment against the
advice of the therapist (i.e., withdrawals), and those who were expelled (i.e., pushouts). The withdrawal group and the pushout group combined were labelled as
the dropout group, which thus can be defined as all patients who prematurely terminated treatment.
Statistical analysis
All analyses were performed using the Statistical Package for the Social Sciences, version 20.0 (IBM, 2011). The variables of interest that had been associated with dropout in previous studies and the behavioural factors that were considered relevant were tested two-sided for differences, with a level of significance of p<.05. A Holm-Bonferroni correction was used to account for the number of characteristics tested.
First, with chi-square tests (categorical variables) and t-tests (continuous variables), the completer group was compared to the dropout group. Subsequently, a binary logistic regression analysis was performed (dependent completer and dropout), and all characteristics were included as independent variables.
The Nagelkerke R-square of the model was used as measure for effect size. To examine assumed differences between withdrawals and pushouts, with chi-square tests (categorical variables) and t-tests (continuous variables), the withdrawal and pushout groups were compared.
83