抄録
Social factors potentially impacting the Japanese divorce rate were analyzed using multiple regression methods for prefectural social sanitary statistics obtained in 2003. The dependent variable was the social factors related to the Japanese divorce rate. The explanatory variables were divided into five categories, which were home ownership, economic status, social life, culture and physical geographical condition. Thirty elements were chosen from those five categories. These variables were confirmed to fit a normal distribution. Subsequently, a stepwise selection method allowed exclusion of multicollinearity from the regression models. Ultimately, multiple regression models containing explanatory variables were obtained. The main divorce promoting factors were found to be unemployment, dual career family, nuclear family, and savings. On the other hands, divorce retardation factors were found to be academic achievement and home ownership. Politics of home ownership was more effects than that of the increasing workfare facility for working youth & women's, which could be automatically compensated by the other politics.
