Integrating Process and Outcome Evaluation
Margaret Blasinsky, Barbara Cohen, Howard Goldman, Kevin Hambrecht, Matthew Johnsen, Wendy Landow, Rob Orwin, and Joseph Sonnefeld
This Toolkit, developed by a team at ROW Sciences, Inc., describes an approach to organizing multi-level or hierarchical data, such as those collected in a multi-site study, to facilitate analysis of data from multiple levels (e.g., characteristics of sites, characteristics of provider or managed care organizations, consumer characteristics, service utilization, and consumer outcomes). The process for constructing a project-level database involves assembling a team, collecting relevant program/site level information from a variety of sources (e.g., contracts, census data, IMHO data) and organizing the information into a database along with more traditional data such as consumer survey and service utilization data. The Toolkit includes an example of analysis using a project-level database developed for a multi-site homelessness service demonstration study funded by the National Institute on Alcohol Abuse and Alcoholism.