Abstract
Increasing numbers of evaluators are cognizant of the need to observe treatment systemat ically. To date, however, little headway has been made toward operationalizing and executing the required effort. In the context of an experiment in criminal justice, a number of "real world" constraints on documenting, monitoring, and measuring treatment imple mentation are discussed. The contract researcher who is hired to conduct an impact evaluation af a black box intervention cannot readily divert the study'sfocus andfunds to an examination of process. By incorporating provisions for gathering implementation data into theirfundingcriteria, evaluationfunders would help to advance the meaningful- ness and social value of outcome studies.

This publication has 19 references indexed in Scilit: