MULTILEVEL COUNT REGRESSION MODELING ON DETERMINANTS OF FERTILITY AMONG REPRODUCTIVE AGED WOMEN IN ETHIOPIAN

Abstract

Background: Fertility is one of the elements in population dynamics that has a significant contribution towards changing population size and structure over time. Following Nigeria, Ethiopia is the second most populous country in Africa. Determining the factors that influence fertility is essential to developing new policies to improve maternal and child health, minimize high rates of population increase fertility in Ethiopia. The main aim of this study was to explore socio economic and demographic factors of fertility in Ethiopia among married women in the Reproductive age group using a count model. Methods: The data was taken from the 2019 EMDHS data, which was gathered representatively across Ethiopia's two city administrations and all administrative areas. A multilevel count model was used to investigate the high risk variables associated with high fertility in Ethiopia with a response variable of the number of living children ever born. Results: According to the findings of the respondents' descriptive study, women have an average of 4.36 living children per mother. The dispersion parameter is examined using the negative binomial regression count model, and the results indicate that it is not significant at the 5% level of significance. The predictor variables, mother’ place of residence, region ,religion, wealth index, age at first birth, ,current age, contraceptive use, marital status, sex of household head, age of household head and age at first birth were found significant determinants at 5% significance level. The initial plot of the expected number of live births versus the different predictors showed regional variations in fertility in Ethiopia. Furthermore, the multilevel analysis demonstrated that, at the 5% level of significance, the variance in the number of living children per mother to be 0.220 with a standard error of 0.1003. Conclusion: In comparison to the Negative Binomial model, it was discovered that the Poisson Regression Model is more appropriate to the data. The single level Poisson regression model's results indicated that Somali regions had the greatest rate of fertility. To minimize Ethiopia’s high fertility rate, it is critical to encourage women to use contraceptives and to wait until marriage in order to raise the age at first birth. Regions with high reproductive capacity should receive extra consideration.