Each individual should continue to be formally and informally assessed at various times throughout his or her incarceration and community supervision. To the extent possible, corrections administrators should adopt instruments for intake that may be modified for use at later decision points, to allow data to be easily compared while eliminating redundancy.

All risk-assessment instruments share some common elements. For example, they all include a list of factors that are "scored" based on some statistical weight that correlates with the risk behavior. In the criminal justice system, most risk-assessment instruments include age as a risk factor (the younger the offender, the higher the risk of recidivism) and some elements related to prior criminal history (the more severe the criminal history, the higher the likelihood of recidivism). Because decision makers must assess for behavior risks at different key decisions points in the criminal justice continuum, however, corrections administrators should not rely on a single instrument or set of instruments for use at these different intervals without validating the instruments for each of these uses. The instruments will need to be adjusted to match the different emphases of the different stages of intake, program planning, release, and supervision. However, measuring the large amount of information that should be collected at each of these points (including dynamic risk factors) in a similar (if not the same) way--using variations on a common instrument--will allow partners to share information more easily and to measure changes over time.

Example: Level of Services Inventory-Revised, Pennsylvania Department of Corrections

In addition to using the PACT instrument at intake (see Recommendation n, above), the Pennsylvania DOC administers the Level of Services Inventory-Revised (LSI-R), to assess both the risk of recidivism and offender needs at a variety of key points: at intake, prior to an individual's review by the Board of Parole, and at regular increments during the supervision period. In 2002, the DOC launched an Assessment Pilot Project during which it tested several tools for possible co-administration with the LSI-R. DOC sought to evaluate the level of an individual's need for intervention in specific problem areas identified as being strongly related to the risk of re-offending. After consulting with research partners and a research psychologist, the DOC selected the Hostile Interpretations Questionnaire (HIQ) and the Criminal Sentiment Scale Modified (CSS-M), and normed them for the Pennsylvania prison population.