| Abstract [eng] |
Criminal risk assessment plays a central role in execution of custodial sentences (Shingler et al., 2018). It aims to target correctional treatment efforts (Scott at al., 2019). Criminal risk is assessed by application of instruments comprised of factors that are robust and significant predictors of future offending (Zara & Farrington, 2013). Various validated assessment tools exist, tailored to different offender groups, including juveniles and adults. However, it is not clear which instruments are more suitable for emerging adults (i.e., from 18 to 24 years old), which, according to Arnett (2002), can be defined as distinct developmental stage. The aim of this study is to evaluate the significance of risk assessment instruments, used in Lithuanian prison service for adults and juveniles, in predicting misconduct among emerging adults serving custodial sentences. The study involved 99 young adult males (M = 20.74, SD = 1.682, aged 18 to 23 years) serving custodial sentences in 4 out of 8 prisons in Lithuania. Criminal risk assessment utilized both the juvenile-oriented START:AV (Viljoen et al., 2014) and the adult-oriented OASys (Home Office, 2002) instruments. Furthermore, violent behaviour, non-violent offenses, substance use, and unauthorized absence were assessed over a three-month period subsequent to the risk evaluation. The results indicated that within the incarcerated emerging adult sample, the START:AV showed significant predictive power for violent behaviour, whereas both the START:AV and the OASys effectively predicted substance use. However, neither risk assessment instrument demonstrated significance in predicting non-violent offenses or unauthorized absence among this group. Given that juvenile and adult risk assessment tools are grounded in distinct factors reflecting different aspects of individuals in transitional age, it's crucial to flexibly consider individual developmental process when selecting risk assessment instruments for predicting misconduct during incarceration. The study was granted by the Lithuanian Research Council No. S-PD-22-101. |
