Statistical assessement of the performance of planting materials on banana varieties in kenya: anapplication on-farm trial

Abstract

On-farm research (OFR) is characterized by its fanner orientation and its interaction with the fanner. On-farm experimentations (OFE's) are used to evaluate the level and consistency of the experimental treatment responses (or interventions) in the face of variations in farmers' circumstances, of climate, soil, product and input markets and fanner management practices. Thus, through active participation by fanners, technologies (or components) are rigorously tested in a more realistic human and environmental background. Unfortunately this may render the design to be complex due to unbalanced ness, high variability between farms, and nonorthogonality among others. Analysis of variance fails. This study addresses this problem with a view of finding a solution to these problems through modelling. Data used in this particular study is an on-farm project involving several fanners within sites (division) on banana disease in Kenya. The project was developed in response to the rapid decline in banana production in Kenya over the last two decades. This decline was due to infestation with Panama, Black and Yellow sigatoka, weevils Inematode complexes and environmental degradation. The common fanner practice of using infected suckers topropagate new banana plant had aggravated the problem further in banana producing areas