Two issues were addressed in the present study. First, the psychometri
c properties of a risk/needs assessment for youth (Youth Level of Serv
ice Inventory (YLSI)) (Andrews, Robinson, and Hoge 1984) were explored
. Second, an effort was made to develop a typology of young offenders.
The sample was comprised of consecutive admissions to one of 6 probat
ion offices in a large urban region and included 256 males with a mean
age of 15.2 years. Statistical analyses showed the YLSI to be psychom
etrically sound and useful in differentiating among offender risk leve
ls. Standardized YLSI subscale scores were subjected to cluster analys
is that revealed a reliable and meaningful five cluster solution. Thes
e clusters were validated against various offence and clinical criteri
on measures.