FERTILITY DIFFERENTIALS: THE CASE OF SOMALIA REGION, EASTEREN ETHIOPIA: APPLICATION OF COUNT REGRESSION MODELS

Abstract

Fertility is defined as the ability to conceive and bears children. It is one of the major components of population growth and age structure change globally fertility. Total fertility rate globally decreasing but in sub Saharan African countries still were increasing. Although there is a significant reduction in Ethiopia over the past decades, but in Somali region of Ethiopia the fertility still high. To assess the determinants of fertility in Somalia region, Eastern Ethiopia by count regression model. A community based-cross sectional study was conducted.1, 391 eligible women were from Somalia region of Ethiopia. Among thus, 1,002 are gave live birth in their life time. This study was undertaken using secondary data from the Ethiopian demographic and health surveys 2016 dataset, which was collected by Central statistical agency of Ethiopia from January 18 to June 27, 2016.data analysis, was done by using count regression model. In this study fertility was treated as count data and determinants of high fertility including the socio economic and demographic variables were assessed using counted data analysis models. The variance (8.195) is higher than mean of child ever born per mother (5.08) and the counted data has no zero value. This is an indication for over dispersion. In this study zero truncated negative binomial better fits the data. The result showed that, the variables like; current age of mothers ,age of mother at first birth, place of residence, number of house hold members, number of died children for both sex, religion, mother education level had significant factors on child ever born fertility. Factors that was high effect on fertility for Somali region, Ethiopia were, mothers first birth is in girl hood age, give birth until natural suspension, their religion, most mothers were uneducated, their residence, intent to increase their household members and high number of dead children in the region were factors to high number of child ever born. By using Zero truncated negative binomial mode.