Methodology/Approach: The paper uses social network analysis (SNA) tools to identify, map and analyze the AISs and the two-stage Heckman selection model. Combining qualitative and quantitative methods allows testing the differential effects of social networks on technology adoption.

Findings: We find weak innovation systems in the study area. Farmers are central in the social networks but have little to no influence in the innovation system perspective. Social network capital plays a critical role in improving adoption of improved farm technology.

Practical implication: Focal farmers are important actors disseminating farm technology. They should be supported by policies to interact more effectively with research and extension agents. Strengthening the role of social networking in the innovation system will enhance adoption of improved farm technology.

Theoretical implications: The inclusion of SNA indicators in the adoption model is a novel approach for quantifying the effect of the innovation system and contributes to understanding the roles of different actors. Further validation of the SNA indicators is needed for a comprehensive analysis of the innovation system.

Originality/value: The innovation system approach is useful for researchers and policy-makers by encouraging them to consider new innovation actors and relationships in technology development and dissemination. Combining quantitative and qualitative methods can help to identify starting points for policies to promote innovations among farmers.  相似文献