A Guide to Longitudinal Program Impact Evaluation

This article serves as a practical guide for implementing longitudinal estimation procedures in program impact evaluation. Throughout, we use as examples the data and objectives of the the Measurement, Learning & Evaluation (MLE) Project of the Urban Reproductive Health Initiative and the MEASURE Evaluation project, both of which are based in the Carolina Population Center at the University of North Carolina at Chapel Hill. The guide first explains the advantages of gathering longitudinal data and then describes the assumptions and methods of analysis for use with longitudinal data. We demonstrate that the applications of these methods can yield the statistically correct estimates of program impact required to answer key evaluation questions, such as those that we have listed for the MLE project. The methods discussion are supplemented with examples from the literature and with empirical examples demonstrated with Stata