Statistical Analysis of Regional Heterogeneity of Marital Dissolution Among Women in Ethiopia, EDHS 2016: Application of Multilevel Logistic Regression Model

Abstract

Divorce is a major life stressor for the individuals involved, with potentially strong negative consequences for the mental and physical health of all members of the family. The aim of this study was to investigate the existence of regional heterogeneity marital dissolution among women in Ethiopia. The study used data from the 2016 Ethiopia Demographic and Health Survey which was a stratified two stage cluster sampling procedure was used. The researcher has been used (n=11405) of all married women from the selected population of study nested within nine regional states and two administrative cities in Ethiopia at time of interview. The Multilevel model were used to explore the major risk factors and regional variations of marital dissolution in Ethiopia using R statistical software. The descriptive result revealed that among eligible married women the proportion of marital dissolution was 9.91%. Among the three multilevel logistic models the random slope model found to be the best description of the data set and to evaluate the within and between regional heterogeneity of marital dissolution. Using this model variables that significantly affect the marital dissolution in Ethiopia were residence, education level of women, work status of women, duration of marriage, number of children, education level of husband and number of unions. The effects of the determinant variables are the same for each region, but the number of children and education level of husband were the two variables which varies within and between in each region. The other important result from this paper is that missing data analysis using appropriate imputation technique was performed to make better inferences.