Page 103 - Age of onset of disruptive behavior of residentially treated adolescents -Sjoukje de Boer

The component ‘substance usage’ was coded as positive when there was no drug use since discharge, if the participant had occasionally used alcohol (not on a regular basis, not more than 1 or 2 drinks per occasion) or sporadically had used cannabis (once or twice a year). Occasional usage of other drugs (e.g., XTC, speed, cocaine, heroin), or regular use of cannabis or alcohol was coded as negative. If there for instance had been a short period of regular use of cannabis or alcohol after discharge, but subsequently it changed for the better resulting in abstinence, this was coded as dubious.
‘Criminal offending’ was coded as positive if the participant reported that he 6 did not commit any crimes or offences. If there had been violations such as traffic offences, fare dodging or fighting, this was coded as dubious. If the participant reported crimes, offences or violations for which prosecution would be justified, this
was coded as negative. It was based on offending that was indicated by the respondent and was rated regardless whether or not persecution actually took place.
Statistical analysis
All analyses were performed using the Statistical Package for the Social Sciences, version 23.0 (IBM, 2015). The inter rater reliability (Cohen’s Kappa) of the components of the general daily functioning scale was determined. To compare the completer and dropout groups on the four (categorized) components of general daily functioning, chi-square tests were performed. Independent-samples t-tests were conducted to compare the completer and dropout groups on the quasi-interval variable general daily functioning at follow-up, and on the continuous variables (e.g., GSI at admission and discharge). First, the completers were compared to the dropouts, and because of the assumed difference between withdrawals and pushouts, both dropout groups were also compared with each other. To compare the completer and dropout groups on (dichotomized) categorical variables (cannabis usage prior to admission, early-onset disruptive behavior prior to age 12, and sex), chi-square tests were performed. To investigate the relationship between the dichotomized categorical variables (cannabis usage prior to admission, early-onset disruptive behavior prior to age 12, and male sex) and general daily functioning at follow-up, independent-samples
101