An application of multivariate methods to evaluate and analyse complementary basic education in Tanzania mainland

Complimentary Basic Education in Tanzania (COBET) is a community-based programme that provides opportunity to over aged children to be enrolled in formal education system. Most of the researches in identifying problems in this programme have employed qualitative techniques. In third world, multivariate analysis techniques have not been applied. Three multivariate analysis techniques were used. These include: Cluster Analysis (CA), Generalized Linear Model (GLM) and Principal Component Analysis (PCA). The study covered 21 Tanzania mainland regions, 424 pupils and 20 key informants. The 21 regions, through cluster analysis were grouped in 4 main clusters. GLM revealed that these clusters varied based on enrollment and dropout. Factors associated with enrollment and dropout were COBET centres, gender, time (years), and cluster of regions. PCA produced three components for dependent variables (72.9%) of total variance and for independent variables (59.8%) of total variance. This study is useful to all stakeholders in understanding factors that influence COBET implementation. Further research should be done on why these centres are no longer active and to develop a tracking system to follow up COBET students.