Using public databases to study relative program impact

Evaluators are under increasing pressure to answer the “compared to what” question when examining the impact of the programs they study. Program contexts and other restraints often make it impossible to study impact using some of our more rigorous methods such as randomized control trials. Alternative methods for studying impact under extreme contextual constraints should be explored and shared for use by others. This article presents a method for studying program impact using existing public datasets as a means for deriving comparison groups and assessing relative impact.