background The Catamaran youth forensic psychiatric hospital Eindhoven offers psychological and psychiatric treatment to youngsters in the ages of 16-24 who have been involved with the criminal justice system and/or pose a risk to themselves or to others. These youngsters often show complex psychiatric problems and their selfesteem is often low. One of the assumptions of the Good Lives Model (Ward & Stewart 2003), a strength-based approach to the treatment of offenders, is that internal conditions, such as selfesteem, are related to rehabilitation and change in behavior.
aim In this presentation, profiles of youngsters self-esteem and behavioral changes were investigated during admission.
methods One hundred youngsters reported about their self-esteem and competences using the Self-Perception Profile for Adolescents (Treffers 2002), these youngsters also reported about their behavioral functioning using the Adult Self Report or Youth Self Report (asr or ysr, Achenbach 2003; 2001) and teacher observations  were available (Teacher Report Form (trf), Achenbach 2003). The profiles of self-esteem were estimated using a latent class analysis on the first measurement during admission. The behavioral changes were described with a growth mixture model, this model estimates the behavioral change for each youngster using the first five measurements of the trf.
results The latent class analysis revealed three groups of youngsters who showed different profiles of self-esteem and competence. Also three different trajectories of behavioral change were found for the trf. The three trajectories were not related to the profiles of self-esteem and competences. However, the profiles of selfesteem showed a stronger relation with risk and promotive factors than the trajectories of behavioral change.
conclusion This study revealed that the competences and self-esteem of youngsters were not related to behavioral changes during admission to a youth forensic psychiatric hospital. However, using the strength-based approach youngsters could be better classified in groups of individuals related to comparable risk and promotive factors.