Beyond Traditional Channels: Post intervention data for Causal Assessment of Social Media's Impact on Smallholder Farmers' Practices in Isingiro district.

The data were collected during a post-intervention follow-up survey for a study designed to evaluate the causal effect of social media on agricultural knowledge and technology adoption among smallholder farmers. The analysis explicitly compares social media to traditional extension channels, investigates the mechanisms driving its influence, and identifies heterogeneous treatment effects across farmer demographics. The findings provide evidence for optimizing the design and scalability of future technology dissemination programs. Methodology:The data were collected under a study that uses a cluster-randomized controlled trial (cRCT) to compare how two different communication channels affect farmers' adoption of improved biofortified bean varieties. The study employed a cluster-randomized design, using sub counties as clusters to create geographical barriers and minimize contamination between groups. Sub counties were randomly assigned to one of three treatment arms: 1. Social Media Extension: Farmers received informational videos via social media. 2. Traditional Extension: Farmers attended in-person training with a local agent. 3. Control Group. The core informational content on bean benefits and seed sources was kept identical across all treatments to isolate the effect of the delivery channel—comparing digital outreach against in-person training. To prevent bias, the study was presented as a general evaluation of delivery methods. The interventions were delivered two weeks before the planting season, followed by an endline survey that measured changes in knowledge and adoption, while also checking for spillover effects. The design ultimately tests whether digital outreach can reduce search costs for improved seed. Following the intervention, endline survey was conducted with the same panel of bean-growing households initially recruited at baseline.