Achieving high-quality impact evaluation design through mixed methods: the case of infrastructure

A good-quality impact evaluation is based on an analysis of the theory of change for the intervention. Analysis of different parts of this causal chain, and the underlying assumptions, necessarily requires use of a variety of research methods. The method should fit the question, not the other way round. The challenge is to genuinely mix these methods rather than conduct parallel studies. The analysis needs to be rooted in a good understanding of context, which may come from anthropology or political science and political economy. Qualitative information sheds light on factors behind programme placement and self-selection. A key contribution from other disciplines is a proper understanding of the nature and distribution of benefits, enabling an impact evaluation design that captures the full range of benefits and socially-mediated impact heterogeneity.