Standardizing outcome data in program evaluation.

Procedures are described for standardizing outcomes in evaluations of service delivery systems. Accounting for the variability of a given device, client gains across services can be expressed in a standard metric of change. Standardized gains take on a common meaning of change regardless of the original scale and variability of the raw data. In return, a wide range of common statistical analyses can be applied with individual client gains as the units of analysis, powerfully delineating influences on outcome variability. Implications and limitations are discussed.