Date of Award


Degree Name

Doctor of Philosophy


Counselor Education and Counseling Psychology

First Advisor

Dr. Eric M. Sauer

Second Advisor

Dr. Mary Z. Anderson

Third Advisor

Dr. Ineke Way


Attachment security, adolescent, residential treatment, change in attachment, residential therapeutic programs, developmental psychology


Adolescents with significant, persistent behavioral and mental health problems are increasingly being treated in private residential treatment programs (RTPs). Recent research at such programs shows that adolescents’ symptoms improve over the course of treatment and that such positive results persist up to a year post discharge. This study attempts to address what is occurring below the symptom level by exploring if attachment security increases as symptoms improve over the course of treatment in private RTPs. The level of attachment security was assessed along the dimensions of attachment avoidance and anxiety as a general construct and according to specific relationships (with mother, father, & therapist). Overall level of symptoms was also assessed. 146 adolescents from four private RTPs participated in the first round of data collection. Those adolescents were then assessed on three more occasions over the span of nine months. The overall number of participants declined over time due to adolescents completing their respective programs.

Using data from the first assessment, between group differences in attachment security were found between adolescents early in their treatment and those late in their treatment. Post-hoc analyses revealed that adolescents in the late stage group had lower scores on general attachment anxiety than those in the early stage group. Similar differences were found between groups with attachment anxiety in regards to mother. With therapist, attachment avoidance was significantly lower for adolescents later in treatment. Growth curve modeling explored how attachment security changed withinsubjects over the nine-month study. For all variables of interest, the linear growth model provided the best fit, indicating that attachment avoidance and anxiety decreased over time in a linear manner. Symptom level declined over time as well. As a predictor, symptom level did not predict the overall growth trajectory of general attachment security but had an effect on the initial scores of both attachment dimensions. Symptom as a predictor also predicted the initial intercept and growth over time for attachment anxiety in regards to mother. Overall findings indicate that attachment security increased over time for adolescents in private residential treatment. This change was paralleled by a decrease in overall symptom level.

Access Setting

Dissertation-Open Access