Complex adoption dynamics of agricultural practices require adaptive mixed-methods research designs

Abstract

Despite heavy promotion of sustainable intensification practices (SIP) among smallholder farmers in sub-Saharan Africa, adoption rates remain low and scaling out has been proven difficult. The dominant approach in adoption research, household surveys with binary adoption variables analyzed through regression models, has produced valuable population-level insights but may give an incomplete picture of how farmers engage with practices on the ground. This paper aims to demonstrate what dimensions of adoption complexity different methods capture and proposes a mixed-methods framework that guides researchers in selecting method combinations suited to the adoption dimension they seek to understand. Drawing on a mixed-methods study conducted in Zomba district, Malawi, combining a district-wide household survey, in-depth farmer interviews, key informant interviews with district officers and extension workers, and farm field visits, we show that different methods will lead to different insights. Farmers adopted and adapted practices in ways that did not appear in survey responses, temporarily dis-adopted and reintroduced practices in cycles that could not be qualified as sustained adoption or dis-adoption. They applied practices unevenly across fields in ways surveys could not detect. Institutional constraints, limited extension reach, project-driven practice promotion, and exclusion from training groups further shaped who could adopt in the first place. A shift toward mixed-methods designs is necessary in identifying adoption evidence that can meaningfully inform food system transformation.