Design, implement, and evaluate farmer field school for digital advisory in maize extension

Thesis Abstract

The rapid digitization of agricultural advisory services presents both opportunities and challenges for maize production in developing economies, where information gaps, low adoption of best practices, and limited access to extension services constrain yield improvements. This study addresses the problem of insufficient, scalable, field-based digital advisory mechanisms by designing, implementing, and evaluating a Farmer Field School (FFS) model integrated with digital tools to enhance maize extension effectiveness. The aim is to determine whether a digitally augmented FFS can improve farmers’ knowledge, adoption of evidence-based management practices, and maize yield and profitability. Specific objectives include (1) developing an FFS curriculum tailored to maize production under digital advisory, (2) implementing the curriculum with a purposive sample of 40 farmer groups (approximately 320 individual farmers) in two agro-ecological zones, (3) evaluating knowledge gain and behavioral change through pre- and post-tests and attitudinal surveys, (4) assessing adoption rates of key agronomic practices (nitrogen management, pest and disease control, hybrid selection, and timely planting) via longitudinal household surveys at 6 and 12 months, and (5) analyzing yield, input efficiency, and profitability impacts using plot-level measurements and farm-level economic analyses. A mixed-methods design combines quasi-experimental and participatory action research elements. The population comprises maize farming households in the selected districts, with sampling conducted in three stages purposive selection of two districts representing contrasting rainfall regimes; stratified random selection of 20 groups per district; and random assignment of 20 groups to the intervention (FFS with digital advisory) and 20 to the control (conventional extension). Data collection instruments include structured knowledge tests, attitudinal Likert scales, farmer practice checklists, farm biometric measurements, and digital usage analytics from the advisory platform. Validity and reliability are addressed through content validity by an expert panel, pilot testing (n=40 farmers), Cronbach’s alpha for attitudinal scales (? ? 0.70), and triangulation across sources. Quantitative data will be analyzed using Difference-in-Differences (DiD) in regression models to estimate treatment effects on knowledge scores, practice adoption rates, yields, and profitability, with controls for baseline characteristics. ANOVA will compare group means across time points, while regression models will incorporate fixed effects for district and participant random effects. Mediation analysis will examine whether knowledge gain mediates the relationship between digital advisory exposure and practice adoption. Qualitative data from focus group discussions and semi-structured interviews will be analyzed thematically to capture experiential insights, barriers, and enablers, with coding conducted in NVivo and triangulated against quantitative findings. Key expected findings include higher knowledge retention and comprehension among participants receiving digital advisory integrated with FFS, significantly increased adoption of recommended agronomic practices, and superior maize yields and net profits compared with the control. The study anticipates that digital tool usage will correlate with timely decision-making, improved input efficiency, and reduced input waste, especially in the zone with greater access to mobile network coverage. Theoretically, the research contributes to Agricultural Extension theory by validating an integrated design that combines experiential learning (Kolb’s Experiential Learning Theory) with diffusion of innovations (Rogers’ Diffusion Theory) and technology acceptance (TAM) within a structured FFS framework. Practically, it offers a scalable, cost-effective model for delivering context-specific digital advisory through community-based learning cycles. The study's contribution to knowledge lies in (a) demonstrated efficacy of digitally augmented FFS in enhancing knowledge transfer, practice adoption, and productivity; (b) empirical evidence on the cost-benefit implications of integrating mobile advisory with field-based learning; (c) a replicable FFS curriculum and implementation protocol adaptable to other crops and contexts; and (d) insights into best practices for ensuring farmer engagement, gender-inclusive participation, and sustainable extension delivery in low-resource settings. Based on the findings, recommendations include policy support for blended extension models, capacity-building for extension agents in digital facilitation, investments in rural connectivity, and adaptation of the curriculum to address climate variability and stress-tolerant maize varieties.