TY - JOUR
T1 - The level of service Inventory-Revised (LSI-R)
T2 - A useful risk assessment measure for Australian offenders?
AU - Hsu, Ching I.
AU - Caputi, Peter
AU - Byrne, Mitchell K.
PY - 2009/7
Y1 - 2009/7
N2 - This study examined the Level of Service Inventory-Revised (LSI-R) with Australian offenders. Normative statistics, offender need characteristics, and the predictive utility of the LSI-R were assessed using a sample of more than 78,000 administrations. No gender differences were apparent on the LSI-R total score, although female offenders scored higher than male offenders on the Finance and Family/Marital subscales. Significant sentence order differences (community, custodial, and a combination of community and custodial) were found, especially on the Criminal History, Education/Employment, and Companions subscales and the LSI-R total score. Bivariate correlations and logistic regression analyses indicated that different subscales were predictive of reoffending by sex and sentence orders. Overall, the assessment instrument indicated modest predictive validity, with varying results for the different offender groups. Female offenders serving combinations of community and custodial sentence orders, however, revealed unexpected findings. Discussions of the results, implications, and recommendations for further research are presented.
AB - This study examined the Level of Service Inventory-Revised (LSI-R) with Australian offenders. Normative statistics, offender need characteristics, and the predictive utility of the LSI-R were assessed using a sample of more than 78,000 administrations. No gender differences were apparent on the LSI-R total score, although female offenders scored higher than male offenders on the Finance and Family/Marital subscales. Significant sentence order differences (community, custodial, and a combination of community and custodial) were found, especially on the Criminal History, Education/Employment, and Companions subscales and the LSI-R total score. Bivariate correlations and logistic regression analyses indicated that different subscales were predictive of reoffending by sex and sentence orders. Overall, the assessment instrument indicated modest predictive validity, with varying results for the different offender groups. Female offenders serving combinations of community and custodial sentence orders, however, revealed unexpected findings. Discussions of the results, implications, and recommendations for further research are presented.
KW - Australian offenders
KW - Criminogenic needs
KW - LSI-R
KW - Normative statistics
KW - Reoffending
KW - Risk assessment
UR - http://www.scopus.com/inward/record.url?scp=67650323691&partnerID=8YFLogxK
U2 - 10.1177/0093854809335409
DO - 10.1177/0093854809335409
M3 - Article
AN - SCOPUS:67650323691
SN - 0093-8548
VL - 36
SP - 728
EP - 740
JO - Criminal Justice and Behavior
JF - Criminal Justice and Behavior
IS - 7
ER -