If there is an interest in analysis of different types of beneficiaries rather than average project impact across beneficiaries, MCC needs to work with the evaluator and project implementers to ensure the program design and supporting evaluation sampling approach allow for analysis of beneficiary subgroups of interest. Under the Peri-Urban Rangeland Project, there was interest in evaluating intensive versus. semi-intensive herders as well as women versus men, but the evaluation was underpowered to capture those effects. The evaluation was able to analyze regional differences. Discussions on sampling frames and related analysis should occur at the outset, including incorporation of these elements into the logic, research questions and evaluation design.
Lesson Learned