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.

本数据集来源于一项旨在评估社交媒体对小农户农业知识与技术采纳因果效应的干预后随访调查。分析明确对比了社交媒体与传统推广渠道，探究其影响机制，并识别不同农户人口统计学特征下的异质性处理效应。研究结果为优化未来技术推广项目的设计与可扩展性提供了实证依据。方法论：本数据集基于一项采用整群随机对照试验（cluster-randomized controlled trial, cRCT）的研究，旨在对比两种不同沟通渠道对农户采纳改良生物强化豆品种的影响。研究采用整群随机设计，以副县（sub counties）为集群构建地理屏障，最大限度减少组间干扰。副县被随机分配至三个处理组之一：1.社交媒体推广组：农户通过社交媒体接收信息视频；2.传统推广组：农户参与当地推广员的面对面培训；3.对照组。所有处理组中关于豆类益处及种子来源的核心信息内容保持一致，以隔离传递渠道的效应——即对比数字推广与面对面培训的差异。为避免偏倚，研究对外呈现为对传递方式的一般性评估。干预措施于种植季前两周实施，随后开展终线调查，测量知识与采纳行为的变化，同时核查溢出效应。该设计最终验证了数字推广是否能降低改良种子的搜索成本。干预后，对基线时招募的同一批豆类种植家庭面板进行了终线调查。