Specification and Estimation of a Biometric Model by Using Logistic Regression for Measuring Child Mortality

Children are the most valuable assets of the Nation. The social and economic development of the Nation has been based on welfare of the children in the Nation. The level of child mortality in a country is not only an indicative of the public health but, it also can be considered as an index of quality of life lived by the people in the country. The various factors influencing child mortality rate are given by: ecological factors, health status variables, vital characteristics, socio-economic characteristics, cultural factors and others. Simple Logistic Regression models can be specified and the estimates of their parameters can be used to measure child mortality in Biostatistics. Generally, Maximum likelihood method of estimation, Non-weighted least squares estimation for the analysis of Bio-categorical data and Discriminant function analysis can be used to estimate the parameters of logistic regression model. In this paper, an attempt has been made by specifying and estimating biometrical models based on Multivariable Logistic Regression Model and Multinomial Logistic Regression Model.