Abstract:
Family planning services enable individuals or couples to determine freely the number and
spacing of their children and to select the means by which it is achieved. Population growth in
Ethiopia is not in parallel with the development of health services and other basic
infrastructures. The objective of this study was to identify the factors that affect utilization of
family planning service among reproductive age women‟s in Ethiopia. Cross-sectional data from
Ethiopian Demographic and Health Survey was used for the analysis. Data was collected by the
Central Statistical Agency from January 18, 2016 through June 27, 2016 and the sampling
technique employed was multistage. A total of 9824 women were considered in this study.
Descriptive analysis, single level, multilevel and Bayesian logistic regression were used for
data analysis using socio-economic, demographic, and proximate variables and utilization of
family planning service as the dependent variable. The results of the study show that, out of a
total of 9824 sampled women 35.83 percent used the family planning services while 64.17
percent did not. The single level, multilevel and Bayesian logistic regression analyses revealed
that the variables that affect the women‟s utilization of family planning service in Ethiopia were
place of residence, age of a woman, religion of a woman, visited by family planning worker,
educational level of women, economic status, knowledge about family planning method, occupation
of women, exposure to mass media, husband education level, husband occupation and number
of having children. The multilevel logistic regression analysis revealed that there was significant
variation with regard to women‟s utilization of family planning services across the regions under
investigation. From the methodological aspect, it was found that random coefficient model is
better compared to the other two models in setting the data well. The results obtained by
applying Bayesian logistic regression analysis show that the standard errors for the variables
incorporated in the model were smaller than the classical logistic regression analysis. This
implies that the Bayesian logistic regression model give a better estimation than the classical
approach.
