Modeling the smoking status of Kenya's males in the presence of missing data.
| dc.contributor.advisor | Ramroop, Shaun. | |
| dc.contributor.advisor | Mwambi, Henry G. | |
| dc.contributor.author | Umugiraneza, Odette. | |
| dc.date.accessioned | 2014-11-21T08:55:29Z | |
| dc.date.available | 2014-11-21T08:55:29Z | |
| dc.date.created | 2014 | |
| dc.date.issued | 2014 | |
| dc.description | M. Sc. University of KwaZulu-Natal, Durban 2014. | en |
| dc.description.abstract | The current research, modeling smoking status in Kenya's males in the presence of missing data has three objectives: The first objective of this study is to identify factors, associated with smoking which will lead to recommendations to the smoking policy in Kenya. The second objective is to apply the appropriate statistical models to model smoking status of Kenya males that incorporates missing data; Logistic regression as well as the generalized linear mixed model are used to model the smoking status. The third objective leads to comparison of the various statistical methods that handle monotone missing data and by their strengths and weaknesses. The following statistical methods for handling missing data are investigated. These are Last Observation Carried Forward (LOCF) and Multiple Imputation (MI) in order to handle the missingness. The missing data will be created by deleting randomly 20% and 30% of the data. The data used is KDHS 2008-2009, the response variable is the smoking status (smoker and non smoker) and the explanatory variables are region, marital status, religion, education, age group of the respondent, wealth index, size of household and access to mass media. | en |
| dc.identifier.uri | http://hdl.handle.net/10413/11627 | |
| dc.language.iso | en_ZA | en |
| dc.subject | Smoking--Statistics. | en |
| dc.subject | Logistic regression analysis. | en |
| dc.subject | Theses--Statistics. | en |
| dc.title | Modeling the smoking status of Kenya's males in the presence of missing data. | en |
| dc.type | Thesis | en |