A statistical study of the experimental techniques used in the research services of the Ministry Of Agriculture
Abstract
In this study. data were re-analysed to
detect possible sources of the high variation
among the observations from completed field
experiments. A method of analysis for block
designs without any particular pattern was
surveyed and various such block designs studied.
Extensive work 8n the method of analysis
(iterative analysis) have been done by Pearcb
(1963, 1975, 19/6), Worthington (1975), Freeman
(1975) and a slightly different version for missing
plot techniques by Preece (1971).
In this dissertation, emphasis has been laid
on the shortcomings of the usual analysis when the
designs are non-orthogonal (Kempthorne, 1978),
existence and development of the iterative analysis
(Corsten, 1958) as an alternative method of
analysis employad in the study and rearrangement
of plots into new blocks.
Data from completed field trials were
subjected to a post-mortem analysis to determine
the block arrangement, plot and block shape,
orientation and size which could have been more
suitable to the experimental sites. It was also
found that besides soil heterogeneity as a major
source of variation, border effects could play an
equally important rcle as a major source of
variation in observations from field experiment.
Citation
Degree Of Master of Science in Agricultural StatisticsPublisher
University of Nairobi Department of Agriculture
Description
This dissertation is submitted in part fulfilment
for the Degree Of Master of Science in Agricultural
Statistics in the University Of Nairobi