Abstract:

This study examined the comparison between Cox and Weibull survival models in estimating the determinants of divorce in Rivers State, Nigeria. Data consisting of Demographic, Socio-economic and Treatment related variables were collected from Judiciary High court for a period of 10 years for the analysis. The Factors estimated were; Age at Marriage of Husband and Wife, Presence of Children, Duration of Marriage, Employment Status of Husband and Wife, Educational Level of Husband and Wife, Number of Counselling Sessions and Court Sittings attended. Cox proportional Hazard (Semi-parametric) and Weibull (Parametric) regression models were compared for a better fit in estimating the determinants of the risk of divorce among couples, using the Akaike Information Criterion (AIC). The result showed that Cox proportional hazards regression model performed better than Weibull regression model with a difference of 44.5 AIC value lower than that of the Weibull model. Hence, Cox PH model revealed that, of all the factors estimated; Employment Status of husband, Presence of children and Duration of marriage had significant effect on the risk of divorce. Specifically, Employment status of husband and Duration of marriage reduced the risk of statutory marriage divorce by 3% and 41% respectfully, while Presence of children in statutory marriage increased the risk of divorce by 72%. The study thereby recommended among others that the husband, who is the head of the family, should strive, struggle, engage and explore legitimate and genuine jobs or businesses to be able to provide the immediate needs of his family, because the marriage of a jobless and idle husband is always at the risk of divorce.