Statistical Analysis of Factors Associated With Early Marriage among Women in Ethiopia Using Multilevel Logistic Regression Model

Abstract

Background: Early marriage is still widely practiced in many parts of the world mainly in developing countries which is particularly dominant in Sub-Saharan African country. Ethiopia has one of the highest rates of early marriage in the world as well as in Sub-Saharan Africa. Objectives: This study is aimed to analyze associated factors of early marriage among women in Ethiopia with consideration of regional variability using multilevel modeling approach. Methodology: The study is made based on the EDHS, 2016 data that has two-stages sampling hierarchical structure, collected for 9825 married women nested within eleven regions with age group 15-49 years. Descriptive statistics, single and multilevel logistic regression model analysis were used to identify determinants of early marriage and its variation across regional states. Results: The results of the study showed that, of 9825 married women considered, 60.8% women were married at early age while 39.2% were married at an age of 18 years and above. The study also identified the major significant factors that affect early marriage among women. As a result, place of residence, religion of respondent, women education attainment, wealth index, husband education attainment, husband occupation status and total number of sibling were found statistically significant at 5% significance level. Furthermore, women who resides in rural area were (OR = 1.239) times more likely to be early married than those lived in urban. The reason behind is that urban women were more educated than those in rural implying that education was an important tools in delaying age at first marriage. The variance of the random component model related to the intercept term is statistically significant; implying the proportion of women got early marriage varies across regions have been accounted by random intercept terms only. Conclusion: Among the three multilevel logistic models the random intercept model found to be the best fitting to the data. Thus, we conclude that those significant factors of women early marriage helps to implement more effective planning policies that target particular units at regional level. Particularly, education and wealth index have positive effect in reducing early marriage. As a result special attention needs to be paid to all regions in order to access education and improve the economic status for young women that help them to reduce early marriage.