Step-wise group testing in the presence of test error
View/ Open
Date
2004Author
Kiboi, David Wamahiu
Type
ThesisLanguage
enMetadata
Show full item recordAbstract
This study is divided into five chapters. Chapter one introduces the basic concepts
in group screening designs. The chapter also has literature review that highlights
the progress made over the years in the area of group screening by other authors.
The scope of work in the study is also introduced with the assumptions that will
be made and notations used.
Group testing without errors in decision is introduced in chapter two. In this
section it is assumed that each group test or individual test yields a correct decision.
The Dorfman and the Sterrett procedures are introduced. The Dorfman procedure
involves testing the group-factor and if found to be defective, all the items are
tested individually. If the group-factor is non defective all the items are passed as
non defective without further testing. The expected number of runs required to
classify all the items in a group-factor as either defective or non defective is also
enumerated ... The Sterrett procedure is introduced in section 2.3. The procedure
is discussed briefly and the expected number of runs required to classify all the
items in the group-factor calculated.
The study on group testing in the presence of test error using the Dorfman
procedure is undertaken in chapter three. A brief description of the test procedure
is given in the introduction. The test procedure in t~~s study involves obtaining two
good readings before one defective reading. The test procedure is undertaken with
the assumption that the group-factor contains one defective item, two defective
v
items and so on until the case when the group is assumed to contain n defective
items. Since each group test will yield a defective group reading, all the items in
the group have to be tested
In chapter four we study the Step-Wise group testing in the presence of test
error. In this chapter, the test procedure involves the factors being randomly
divided into 'g' groups called group-factors. These group-factors are then tested
for significance. If a group factor is declared defective in the first test, testing of
the items is done individually until the first defective item is found. The remaining
items in the group are pooled together and a group test performed. If the pooled
group test is significant, then items are tested one by one until the defective item
in the pool is reached and the remaining items pooled again. This procedure is
continued until the group test yields two good readings before a defective reading
is obtained. This procedure will be called the modified step-wise group-screenin?
procedure herein referred to as the procedure. If the group test yields two good ,--
readings before a defective reading is obtained then the group is passed as non ,:
defective and all the items in it passed as good without further inspection.
In chapter five the results from simulated data are obtained and tabulated. The
results obtained are compared to the results obtained by Patel and Manene(11).
The results are discussed also in this chapter.
Finally, we have the appendix with the testing s2heme in the form of a diagram
and bibliography detailing the source of reference materials used for this study.
Citation
Masters of Science in Mathematical StatisticsSponsorhip
University of NairobiPublisher
Department of Mathematics University of Nairobi