Effects of CBT Treatment On Attachment Styles Of Incarcerated Sex Offenders

Background and Purpose: The development of a definitive theory of the etiology of sexual offending remains elusive. In an attempt to understand the origins of such behaviors, researchers and clinicians have applied multiple explanatory theories, including attachment theory. Studies on attachment theory have demonstrated that when compared to those children with secure attachments, children who have insecure attachments score lower on empathy levels, impulse control, affective regulation, social skills, and relational capacities, referred to in criminology as criminogenic needs. Deficits in these domains have also been noted in sexual offenders. As such, many sex offender treatment programs target these exact areas to reduce risk levels. Despite these parallels, little research exists evaluating the connection between attachment and sexual offending, and even less on whether treatment has any effect on attachment styles in sexual offenders. The purpose of this study was to test whether attachment styles change over the course of an intense residential sexual offending treatment program for incarcerated adult male sex offenders.

Methods: A quasi-experimental design was used to examine the two research hypotheses: 1) incarcerated male sexual offenders who participated in treatment will exhibit a statistically significant difference on attachment measures post-treatment compared to pre-treatment; and 2) treatment completers will demonstrate statistically significant scores on the attachment measures at post-treatment when compared to the waitlist group. Treatment completers participated in the Sexual Offender Adult Rehabilitation Program (SOAR), which is a 40-hour/week, five-month cognitive-behavioral treatment program administered by the North Carolina Department of Public Safety. The final sample of 34 male sex offenders (treatment n = 16, waitlist n = 18) was drawn from the same medium security level prison. All of the participants had volunteered to participate in SOAR. To measure attachment styles, each participant completed the Relationship Scales Questionnaire (RSQ) and the Adult Attachment Scale (AAS) as well as other psychosocial measures not discussed in this study, at pre-treatment and five months later at post-treatment.

Results: The null hypothesis was rejected for both research questions. 1) Paired sample t-test results indicated that on the RSQ participants had significant increases in their level of secure attachment and significant decreases in their level of fearful attachment with a 4-factor attachment measure. Compressed to 2-factors, participants in the treatment group demonstrated a significant decrease in avoidance and anxious attachment measures. 2)  ANCOVA results indicated that on the RSQ participants in the treatment group had significant increases in their level of secure attachment with a 4-factor attachment measure, compared to the waitlist group. Compressed to 2-factors, participants in the treatment group demonstrated a significant decrease in anxious attachment levels at post-test, compared to the waitlist group.  Interestingly, none of the measures on the AAS were significant.

Conclusions: Although this study had a small sample, the findings support the thesis that intensive cognitive behavioral programs targeting criminogenic needs impact general attachment styles. Attachment theory should be explored further to determine how changes in attachment style influence risk levels and recidivism in sexual offenders, even in the context of a CBT-based intervention.