Mobile-Enabled Extension Service for Smallholder Crop Diversification

Thesis Abstract

The study addresses the persistent gap in access to timely, location-specific agricultural information for smallholder farmers and the consequent constraints on crop diversification opportunities in rural markets. Despite policy emphasis on leveraging information and communication technologies (ICTs) for agricultural development, extension services remain fragmented, inefficient, and unevenly distributed, limiting farmers’ ability to adopt diversified cropping systems that enhance resilience and productivity. The aim is to evaluate the effectiveness of a mobile-enabled extension service (MES) designed to support smallholder farmers in diversifying crops, with specific objectives to (i) assess uptake and usage patterns of the MES among smallholders, (ii) determine the impact of MES on knowledge acquisition about diversified crops and on production diversification decisions, (iii) examine the role of farmer characteristics and contextual factors in mediating MES effectiveness, and (iv) identify barriers and enablers to sustained use and scale-up of MES. The study adopts a quasi-experimental design with a mixed-methods approach, combining quantitative and qualitative data to capture both measurable outcomes and user experiences. The population comprises smallholder farmers across three distinct agro-ecological zones. A stratified random sample of 420 farmers is selected, with 210 assigned to the treatment group receiving MES interventions and 210 to a control group receiving standard extension services, over a 12-month period. Data collection instruments include a structured survey capturing knowledge, attitudes, and practices related to crop diversification, transaction logs from the MES, and field-level agronomic measurements (yields, input usage, and diversification indices). In-depth interviews and focus group discussions (n=30 and n=12 respectively) supplement quantitative data to explore perceived utility, trust, and contextual constraints. Validity and reliability are ensured through pilot testing, item analysis, and Cronbach’s alpha assessments for survey scales, along with triangulation across data sources. Quantitative data are analyzed using difference-in-differences (DiD) estimation to identify causal effects of MES on diversification outcomes, accompanied by multivariate regression analyses controlling for baseline characteristics. Mediation analyses examine pathways from information exposure to behavioral change, while logistic regression models assess the likelihood of adopting specific diversified crops. Qualitative data are analyzed thematically using a framework approach to extract core themes related to usability, trust, and contextual fit, with triangulated coding conducted by two independent researchers. Additionally, a technology acceptance model (TAM) is adapted to interpret user adoption patterns, and insights are aligned with the Diffusion of Innovations theory to explain uptake dynamics. Key expected findings include significant improvements in knowledge scores related to diversified crops, higher diversification indices among MES users, and greater adoption of climate-resilient and market-oriented crops compared with controls. The study anticipates differential effects across gender, literacy levels, and proximity to mobile networks, highlighting equity considerations in MES deployment. The contribution to knowledge lies in empirically validating a scalable, ICT-driven extension framework that integrates behaviorally informed design with agronomic outcomes, advancing theories of digital extension adoption in low-resource settings and providing a replicable model for policy and practice. The research also offers practical guidance on content localization, offline capabilities, data privacy, data-driven recommendations, and institutional partnerships essential for sustaining MES at scale. The main conclusion is expected to be that a well-implemented MES can enhance knowledge transfer, decision-making, and crop diversification while recognizing the importance of supportive enabling environments, including training, network reliability, and trusted extension personnel. Recommendations include iterative co-design with farmer communities, investment in network infrastructure and digital literacy, alignment with agricultural input suppliers and input credit schemes, and policy measures to institutionalize MES within national extension architectures. The study anticipates informing future research on long-term impacts on income diversification, risk reduction, and agro-biodiversity, with a proposed longitudinal follow-up to assess sustained effects and scalability across varied agro-ecologies.