The use of action variables in determining recommendation domains: Grouping senegalese farmers for research and extension

The Ziguinchor Region of Senegal is considered the breadbasket of the nation, yet agricultural production has consistently failed to meet expectations and needs. The reason given for the deficit in production is the failure of researchers and development agents in providing appropriate technologies to targeted farmers with similar priorities, potentials and constraints. This group of farmers form recommendation domains (RDs).

In this paper, cluster analysis was used to classify farm family units into RDs using endogenous and exogenous variables to the farming system. Canonical-discriminant analysis was used to identify the factors influencing homogeneous groups of farmers in attaining food self-sufficiency. This process of isolating RDs is considered appropriate in providing researchers with an effective tool in determining the factors which affect farmers' productivity, and hence articulating research and extension efforts to meet farmers' needs. Information gathered through this method can then be more readily translated into a policy framework.