Participatory Digital Extension for Smallholder Maize Farmers, Tanzania

Thesis Abstract

Smallholder maize farming in Tanzania faces persistent gaps in timely access to agronomic information, farmer training, and market and input services, limiting productivity and resilience amid climate variability. This study investigates how participatory digital extension platforms can enhance knowledge exchange, decision support, and adoption of improved practices among smallholder maize producers. The aim is to design, implement, and evaluate a participatory digital extension (PDE) intervention that integrates mobile-enabled agronomic advisory services, farmer field schools (FFS) via digital media, and community feedback loops to co-create context-appropriate extension content. Specific objectives are (1) to assess baseline extensional knowledge, technology adoption, and maize yield determinants; (2) to co-design and deploy a PDE ecosystem featuring interactive advisory modules, offline-capable mobile apps, voice-based support, and farmer-to-farmer learning networks, implemented in Kilosa and Mvomero districts; (3) to evaluate changes in knowledge uptake, adoption rates of recommended practices (soil fertility management, improved seeds, pest and disease management, and climate-smart techniques), and maize productivity over two cropping seasons; (4) to examine governance, inclusivity, and equity in access to digital extension across gender, age, and socio-economic strata; (5) to generate policy and practice recommendations for scaling PDE in Tanzania. A mixed-methods research design will be employed, combining quasi-experimental evaluation with participatory action research. The study population comprises 2,400 smallholder maize farming households, with a purposive subsample of 400 households allocated to intervention and 400 matched controls within the same districts for impact assessment. Data collection instruments include structured household surveys, farmer diaries, focus group discussions, key informant interviews, mobile app usage logs, and agronomic field measurements. Validity and reliability will be ensured through pre-tested instruments, triangulation across data sources, and measurement of internal consistency (Cronbach’s alpha >0.7) for attitudinal scales. Descriptive statistics, t-tests, and chi-square tests will describe baseline characteristics; difference-in-differences (DiD) analysis and propensity score matching (PSM) will estimate causal impacts of PDE on knowledge acquisition, practice adoption, and yield outcomes. Multivariate regression models will identify determinants of adoption and productivity gains, while structural equation modeling (SEM) will test the theoretical pathways linking information access, decision support, farmer empowerment, and performance. The study will also apply thematic analysis to qualitative data to extract insights on user experiences, inclusivity, and barriers to digital extension adoption. The theoretical framework draws on the Diffusion of Innovations (Rogers) and the Technology Acceptance Model, complemented by a Capability Approach to assess empowerment and agency in information access and decision-making. Expected findings include (i) higher knowledge scores and more rapid adoption of climate-smart and soil fertility-enhancing practices among PDE participants; (ii) statistically significant increases in maize yield and gross margin relative to controls, with DiD estimates indicating a 12–18% yield gain and 15–20% efficiency improvements; (iii) enhanced gender-inclusive participation and marginalized groups’ access to digital advisory services; (iv) positive associations between platform usability, perceived usefulness, and sustained engagement; (v) critical facilitators and constraints in scaling PDE, including connectivity, literacy, and local content relevance. The study contributes to knowledge by integrating participatory design with digital extension to operationalize inclusive, context-specific advisory services in smallholder farming systems, and by providing robust empirical estimates of PDE's impact on productivity and livelihoods in Tanzania. The main conclusion is that participatory digital extension can significantly improve knowledge dissemination, technology uptake, and maize productivity when co-designed with farmers, ensure offline and vernacular content, and incorporate continuous feedback loops for local adaptation. Recommendations include scale-up strategies with public–private partnerships, investment in rural connectivity and digital literacy, gender-responsive design, and policy support for data privacy and farmer-owned data governance.