Summary: | 碩士 === 世新大學 === 行政管理學系 === 105 === This study uses traditional regression and Geographically Weighted Regression (GWR) to analyze the factors related to Fertility Rate over 2014-2016.First, through the literature review, this study generalizes the factors related to fertility rate as following: (1) higher education ratio of women (2) ratio of low-income population (3) ratio of obstetrics and gynecology medical resources (4) marriage rate (5) ratio of aboriginal population (6) median household income. Secondly, in terms of the results from regression analysis, this study finds that Geographically Weighted Regression Model is better than the traditional regression model. In addition, according to the map, the results show that the association between the dependent variable and independent variables has regional differences. And these variables in the results of the analysis showing the spatial heterogeneity phenomenon are as following: (1) ratio of low-income population (2) marriage rate (3) median household income. These are the phenomenons which cannot be discovered by traditional regression models. Therefore, this study suggests that the differences in the various regions of Taiwan must be taken into account. The consideration and planning which ingnore the spatial heterogeneity, may not be able to effectively improve fertility in some areas, and may cause policy failure.
|