Table 2.

GWR model estimation Results

	Mean		Median		Min		Max
	Year 2004	Year 2008	Year 2004	Year 2008	Year 2004	Year 2008	Year 2004	Year 2008
PM2.5	0.0047	0.0096	0.0019	0.0110	−0.0040	− 0.0281	0.0274	0.0391
lngdp	0.1032	0.1559	0.1399	0.1823	−1.0014	−1.1763	0.5566	0.6384
lncost	0.4422	1.2551	0.1921	0.6807	−1.3661	−0.5790	2.0427	6.5961
lnmedi	−0.1449	0.2783	−0.1103	0.1668	−0.8896	−0.6085	0.1797	1.9552
popu	0.0002	−0.0001	−0.0011	0.0009	−0.0052	− 0.0086	0.0018	0.0032
intercept	−2.0213	−1.3147	−0.4548	−5.6989	− 21.0838	−7.9364	4.7124	5.2347
R2	0.5660	0.6774	0.5660	0.6774	0.5660	0.6774	0.5660	0.6774
Adj R2	0.4395	0.5833	0.4395	0.5833	0.4395	0.5833	0.4395	0.5833

lngdp: per capita gross domestic product; lncost: per capita medical expenses; lnmedi: number of provincial general hospitals; popu: population density

Geographically weighted regression could report estimated coefficients for each variable in each province. Therefore, there will be 31 estimated coefficients for each variable, corresponding to 31 provinces. In order to compare them between different period, the average, median, minimum, and maximum values of the 31 estimated coefficients were calculated and reported in Table 2