Use Of Logistic Regression Model To Determine The Key Contaminants Of Water In Yatta District

Abstract

Over one sixth of the world's population lacks safe drinking-water sources. The water crisis in Kenya is due not only to the wave of droughts but also to poor management of the water sources, pollution of water supplies by untreated sewage and a huge population explosion. In the study, the main contaminants that could make drinking-water unsafe in Yatta district are determined. The data used to analyse the contamination levels of water were collected during a survey conducted by Kenyatta National Hospital (KNH), Department of Public Health (DPH). The survey covered water sources in Kithimani area of Yatta district between January and December 2009. The survey was based on physio-chemical parameters on the degradation in the level of the quality of drinking-water. The parameters tested were pH, color, sulphates, lead, nitrate and turbidity. Water samples from different sources within the area were collected including both surface and ground water. Logistic regression model was used to determine the key contaminants of the water. The study shows that there was significant level of contamination of the drinking-water on most of the sources. pH was noted to be the predictor which was the most significant on the non-safety of the water.