Learning to use sas to analyze biomedical data using data collected for: urban-rural differences in lung function in Kenyan school children

Abstract

The overall goal of this thesis was to learn how to analyze biomedical data using SAS, a software system for data analysis. In order to achieve this I joined a SAS training seminar held on July 18-22, 2003, at the Centres for Disease Control and Prevention (CDC) located within the KEMRI campus. I was one of 6 participants; all other 5 were from the CDC Global AIDS Program (GAP) Data Section. In order to learn how to use SAS, we used Lung Function data collected as part Kenya Asthma Surveys of Childhood Asthma. The main goal of these surveys was to examine the impact of urbanization on the occurrence of asthma in Kenyan Schoolchildren. The specific objectives for this thesis were to learn how to perform univariate, bivariate and multivariate', analysis for continuous and discrete variables. Specifically, I set out to learn how to handle continuously measured variables (Lung function FYC and FEY 1] ~nmultivariate regression analysis, examining whether 4 assumptions of linearity were satisfied or not and to come up with an urban vs. rural comparison in age and size adjusted lung function level.