Page 104 - Age of onset of disruptive behavior of residentially treated adolescents -Sjoukje de Boer

t-tests were used. A level of significance of p<.05 (two-sided) was chosen, with a Holm- Bonferroni correction to account for the number of characteristics tested.
Univariate regression analysis was conducted with discharge status (completer versus dropout), as independent variable, and the general daily functioning score as dependent variable. Preliminary analyses were conducted to ensure no violation of the assumptions of normality, linearity, multicollinearity and homoscedasticity. Subsequently, to control for the possible effect of cannabis usage prior to treatment, early-onset disruptive behavior, and male sex, these dichotomized variables were each included as independent variables in separate regression analyses. Finally, a multiple regression analysis was performed, in which the discharge status (completer versus dropout), was included as independent variable, and the general daily functioning score as dependent variable, controlling for the possible joint effect of cannabis usage prior to treatment, early-onset disruptive behavior, and male sex. The Nagelkerke R- square of the model was used as measure for effect size.
Results
General daily functioning
Scores on general daily functioning were normally distributed (Skewness = -.58 Kurtosis = -.73), with a mean of 5.3 (range 0-8; SD=2.4). The inter rater reliability (Cohen’s Kappa) of the four components by two research assistants on a random subset of the sample (n=15) showed that the inter rater reliability of the ‘stability of living condition’ was moderate (Cohen’s Kappa .55), of ‘daytime activities’ excellent (Cohen’s Kappa .88), and those of ‘substance usage’ (Cohen’s Kappa .64), ‘criminal offending’ (Cohen’s Kappa .75), as well as the sum of the components, ‘general daily functioning’ (Cohen’s Kappa .75), good (Fleiss, 1981).
Completion and dropout
About one third of the 196 participants had dropped out of treatment (34.2%). Completers had a higher score on the general daily functioning scale at follow-up (see Table 2), indicating that in the eighteen months after discharge they functioned better than those who had dropped out.
102