Logistic Regression Analysis of Marriage Instability in Nigeria

Abstract

This study is aimed at fitting binary logistic model that will describe the pattern of marriage instability in Nigeria in order to identify the factors that trigger marriage instability in Nigeria. The data used for this paper are secondary data from the Nigeria Demographic and Health Surveys (NDHS) reports for 2008. These are national population based surveys conducted across the six geo-political zones in Nigeria. The survey included registered married women across the country.

Binary logistic regression was used to analyze the data on number of unions extracted from National Demographic and Health Survey (NDHS), 2008. Based on the nature of the NDHS data, 24 independent variables (predictors), all categorical were suspected to influence the response variable (number of unions). From the six geopolitical zones, our result shows that the risk of having more than one union in north east zone is 123.6% more than the risk of having more than one union in north central. Also the risk of having more than one union in North West and south west are 92.3% and 26.6% respectively more than the north central. And the risk of having more than one union in south east is 38.9% less than the risk of having more than one union in the north central revealing  that instability marriage is very common in north east, north west and south west in Nigeria.