Dealing with missing data in demographic and health surveys
Abstract
This project deals with the problem of missing data in the
Tuunganepillars of Hope -PWP (2008) survey carried out by impact Research and
Development organization (IMPACT RDO)in Kisumu, Kenya. The emphasis of this
project ison Expectation Maximization (EM) algorithm technique as a method of
handlingdata problem in multivariate analysis. The primary objective of the 2008
PWP project was specifically to enhance supportive and informed home
environments of HIV-infected youths and promote safe sexual behaviors and
generallyto provide up to date information for policymakers, planners,
researchersand program managers, which would allow guidance in the planning,
implementation, monitoring and evaluation of health programs in Kenya.
However,when faced with missing values, researchers or practitioners in research
usuallyproceeds as if the omitted cases had never really been observed.
Citation
M.Sc (Mathematical Statistics) ThesisSponsorhip
University of NairobiPublisher
School of Mathematics, University of Nairobi
Description
Master of Scince Thesis