Application of Geographically Weighted Regression Model(GWR) in the Economics of Health and Health Care: Based on the In-troduction of Arc GIS 10.4
- VernacularTitle:地理加权回归模型在卫生经济领域的应用举例:基于Arc GIS 10.4软件介绍
- Author:
Tao SHI
1
,
2
;
Shi-Xue LI
Author Information
- Keywords: geographically weighted regession model; basic public health service level; influencing factor; spatial auto-correlation; OLS
- From: Chinese Health Economics 2018;37(10):10-14
- CountryChina
- Language:Chinese
- Abstract: Objective: It analyzed the current situation and influencing factors of provincial essential public health service level index(PHI) in 2016. The ordinary least square(OLS) and the geographically weighted regression model(GWR) were constructed respectively. The advan-tages of GWR in the field of health economics were compared. Methods: Moran’s I was used to analyze spatial auto-correlation and hetero-geneity of PHI. OLS and GWR models were constructed to estimate the main influencing factors and their directions. Results: There were a positive spatial auto-correlation in provincial PHI and a weak agglomeration pattern in geographical distribution; the level of economic devel-opment and government funding, population size, population structure and population urbanization level were the main influencing factors of the PHI. There was spatial heterogeneity in provincial regression indexes estimated by GWR. Conclusion: When dealing with geographically related issues in the health economics, the GWR model was better than the OLS with a higher degree of goodness of fit.