What kind of methodological standards are expected from the evaluation design?
3ie supports impact evaluations that adhere to agreed-upon methodological standards for addressing the “attribution challenge” – e.g. establishing cause and effect between programmatic activities and specified outcomes. In particular, evaluation designs must be capable of addressing: a) confounding factors; b) selection bias; c) spillover effects; d) contamination of control groups; and e) impact heterogeneity by intervention, beneficiary type and context.
