The best way to ensure that the intake process is efficient and effective is to use tools that have been validated by research. Validity applies to both face value (whether the instrument or process used makes sense to those who use it) and predictive validity (whether the instrument or process demonstrates a capacity to predict and measure risks and needs based on statistical association). Validation should include a review of how decision makers have used the instrument to guide their clinical decisions. (See Research Highlights, above, for more on the meaning of validation.)

Initial validation should be done by the instrument's developer. Correctional administrators or research staff should further gauge the validity of the instrument for their own usage, either through the agency's own research department, or in conjunction with the instrument's developer. Statistical data should be collected that compares the results of instrument identifications to outcome information (e.g., re-arrests, recidivism, program success or failure) for all screened individuals.

Instruments must also be "normed" for the population which will use them. That is, they must be calibrated to ensure that results for the new population, such as the population of male prisoners in a particular facility, are as valid as they are for the population for which the instruments were developed. The validation and norming procedure should also determine the extent to which the instruments employed are sensitive to cultural and gender variations. Since the demographics and other characteristics of the population entering the correctional facility will change over time, corrections administrators should ensure that the validation is updated and improved as much as possible given available resources.

Example: Pennsylvania Additive Classification Tool, Pennsylvania Department of Corrections

In 1991, the Pennsylvania Department of Corrections (DOC) used a grant from the National Institute of Corrections to develop the Pennsylvania Additive Classification Tool (PACT), an instrument designed to measure an individual's risk level during the period of incarceration and to establish custody levels. PACT was developed by an advisory team and a consultant from the National Council on Crime and Delinquency and normed against a 900-person sample of the Pennsylvania state prison population. During a recent revalidation of pact, the DOC found that the instrument had strong predictive validity and required only minor adjustments.