Application of Design of Experiments in Population Analysis

Abstract

One of the components of population analysis is the study of differential fertility among different groups of mothers classified by child mortality level, regional variation and or socioeconomic status. Among different socioeconomic group of mothers, the levels of fertility or the levels of child mortality are different. The homogeneity of differentials in fertility levels or in child mortality levels can be investigated using the technique of combined analysis of groups of experiments under both homogeneous and heterogeneous conditions. In the present paper, homogeneity of fertility levels by child mortality levels and homogeneity of child mortality levels by number of ever born children recorded from North-Eastern Libya are investigated. The data show that child mortality levels do not vary due to the variation in the number of ever born children of mothers. The level of occupation of mothers changes with the change in their level of education. The study of level of fertility variation according to changes in level of occupation nested within the levels of education is of interest. Again, fertility variation is influenced by duration of marriage, level of education of husband, duration of breastfeeding period etc. Thus, fertility variation should be studied after eliminating the impacts of the above said variables. This can be done using the technique of covariance analysis in nested classification. The method of analysis is explained with real world data.