Welcome to the International Institute of Tropical Agriculture Research Repository
What would you like to view today?
Using a linear discriminant analysis approach of baseline conditions to develop household categories in the Sudan Savanna ( KKM PLS SSA CP ), Nigeria
MetadataShow full item record
To address the problem of the continued deterioration of livelihood and food security in the sub Saharan Africa, the Forum for Agricultural Research in Africa (FARA) through the sub Saharan Africa Challenge Programme (SSA CP) is implementing the Integrated Agricultural Research for Development (IAR4D). The IAR4D is a multistakeholder agricultural research approach which is currently being implemented at Pilot Learning Sites (PLSs) in three regions of Africa: (I) the Kano-Katsina-Maradi (KKM) PLS in West Africa, (2) the Lake Kivu (LK) PLS in East and Central Africa and (3) the Zimbabwe-Malawi-Mozambique (ZMM) PLS in Southern Africa. The objective of this paper was to employ some baseline data of the Sudan Savanna Task Force of the KKM PLS in West Africa, in a linear discriminant analysis to investigate some of the factors that characterised the farmers based on some starting conditions. The study was also to show whether the farmers that have been baselined have common characteristics that can hypothetically separate them on the basis of belonging to three distinct groups for the implementation of the IAR4D. The sampled respondents were initially classified into three groups of baseline farmers. The grouping was done on the basis of whether the farmers are IAR4D (intervention) or conventional (ARD) or clean sites farmers. This is necessary for the end line survey and for the impact evaluation of the programme. Data on a sub-sample of 300 baselined respondents were used for analysis (92-IAR4D/intervention farmers, 96-ARD/conventional farmers and 112-clean farmers). Results indicated an overall rate of 99% of farmers correctly classified into their respective sites. A number of indicative baseline variables (about 67% of the hypothesized variables) which can be regarded as those which distinguish farmers into those which predictably belong to IAR4D/intervention farmers, ARD/conventional farmers and clean farmers were identified to be significantly important. The different villages chosen for the program evaluation are also correctly identified within their groups. Therefore, three distinct categories of villages are available for evaluating the programme impact.