Hi, For the replication assignment, my interest is to identify an evaluation of a health or development intervention, either the design or analysis of which was sub-optimal and possibly adjust the estimate of the treatment effect. There's a lot of recent literature on evaluations and new approaches to working with non-randomized or observational data. There are new estimators for obtaining the treatment effect where unobserved factors otherwise hinder inference and new software for improving balance and assessing overlap in the sample. These new methods may allow us in replication to add significant value to evaluations published previously in the absence of these tools. Let me know if anyone has any interest. Thanks. Andy Stokes