Meta-analysis as a method of multi-site evaluation: An example from international development

Interest in multi-site evaluation has seemingly grown, with a now nascent body of literature. Even so, most of the literature on multi-site evaluation has come from writings about place-, group-, and cluster-randomized controlled trials. Guidance on how to execute uncontrolled multi-site evaluations is sparse and usually treats sites as homogeneous. In this paper, metaanalysis is used to demonstrate the method's potential for multi-site evaluation using Heifer International sites in Albania, Nepal, and Uganda as an exemplar, with an emphasis on nutritional outcomes for site participants.