Lessons from using cluster-randomized evaluations to build evidence on large-scale nutrition behavior change interventions

The recent Nobel Prize in Economics for the use of experimental research to identify solutions to a range of development issues resonates with our work in nutrition. For over a decade, our research team has worked with a global nutrition social and behavior change initiative and used cluster-randomized evaluations, with other methods, to generate lessons about nutrition behaviour change at scale in three countries: Bangladesh, Vietnam, and Ethiopia. We also tested adaptations in other countries. We learned that large-scale behavior change interventions delivered through diverse platforms (government health systems, community-based platforms, and mass media) had substantial impacts but that these impacts differ by context. A body of evidence, based on these evaluations, now informs approaches to shaping nutrition behaviors around the world. Working closely with implementers, sharing research findings and lessons in many forums, and publishing widely, Alive & Thrive has benefited millions of women and children and their communities and influenced millions of dollars of spending on nutrition programs. We conclude that carefully done collaborative program evaluations that use randomized controlled trials together with other methods can support effective learning about solutions, even those that operate at scale.