Balancing Efficiency and Accuracy: Guidelines for the Psychological Assessment of Extended Supervision Orders for Child Sex Offenders in New Zealand

Abstract

The New Zealand Parole (Extended Supervision) Amendment Act (2004) legislated that individuals with specified sexual offences, principally child sex offenders, could be subject to an extended supervision order (ESO) based, in part, on completion of a “psychological health assessment” (PHA). However, after several years of these assessments, fewer than half of PHAs have resulted in ESOs, suggesting that the PHA referral threshold may be set too low, resulting in a large number of false positives. This research therefore aimed to explore whether there were more efficient means to determine appropriate PHA referral guidelines. Participants were 182 offenders who had been assessed for an ESO between 2004 and 2008. The method included reviewing participants’ PHA reports, static risk scores and criminal conviction histories. Discriminant function analysis revealed that the factors most predictive of an ESO being recommended were: the presence of a higher number of dynamic risk factors related to sexual offending, older age and a higher number of historical sexual offence victims. Based on the current findings, further research is recommended to determine how we can reduce the number of unnecessary and aversive assessments through better screening from file scored dynamic pre-assessment checks.

Key words:

• child sex offender
• dynamic risk assessment
• extended supervision
• STABLE-2007

Notes

1. There is also provision in the ESO legislation to include offences against intellectually disabled adults.

2. Item 5 on the ASRS (prior sentences for non-sexual violence) also taps into antisocial behaviour and attitudes. Two items on this measure specifically assess for general antisociality (1) index non-sexual violence and (2) prior non-sexual violence.

3. Of note, a small number of the men in the current study also had had formal assessments of psychopathy conducted on them using the Psychopathy Checklist Shortened Version (PCL:SV, Hart, Cox, & Hare, 1995). However, due to the small number that had undergone this assessment, this information was not formally employed in the current study.