WLS输出结果怎么看

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

如图，下面是我做得WLS结果，1.请问下面的系数是我原来的变量前的系数，还是经过变换后的新变量的系数2.下面的weighted statistic 和unweighted statistic有什么去边，看不懂

Dependent Variable: RP
Method: Least Squares
Date: 05/03/12   Time: 18:35
Sample: 1 31
Included observations:  31
Weighting series:  1/ABS(R)
Variable	Coefficient	Std. Error	t-Statistic	Prob.
C	-2607.395	838.1469	-3.110904	0.0045
GDP	0.063558	0.030531	2.081781	0.0473
UL	71.57191	25.82114	2.771834	0.0102
I	9666.710	2183.345	4.427477	0.0002
D1	1260.079	318.1239	3.960969	0.0005
	Weighted Statistics
R-squared	0.961856	Mean dependent var		4695.806
Adjusted R-squared	0.955988	S.D. dependent var		8563.265
S.E. of regression	526.9428	Akaike info criterion		15.51875
Sum squared resid	7219387.	Schwarz criterion		15.75004
Log likelihood	-235.5406	Hannan-Quinn criter.		15.59415
F-statistic	163.9081	Durbin-Watson stat		1.758911
Prob(F-statistic)	0.000000
	Unweighted Statistics
R-squared	0.755810	Mean dependent var		5138.161
Adjusted R-squared	0.718242	S.D. dependent var		3451.770
S.E. of regression	1832.230	Sum squared resid		87283705
Durbin-Watson stat	1.673788

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：输出结果 Wls observations observation coefficient Error

一般是关注系数t检验

基本的系数检验在0.05的置信度下通过检验，R方较高拟合较好，模型总体的F检验也通过
你说的区别是加权和不加权的区别