spss和Eviews 线性回归结果不一致 为什么?
同样的因变量 和自变量
做多元线性回归结果却不一致?
结果输出如下:
spss结果:
Coefficients(a)
Model Unstandardized Coefficients Standardized Coefficients Collinearity Statistics
B Std. Error Beta t Sig. Tolerance VIF
1 (Constant) -397076.344 104501.812 -3.800 .003
第一产业比重 -7173.603 1299.198 -.365 -5.522 .000 .232 4.307
第二产业比重 13936.633 2415.515 .209 5.770 .000 .778 1.285
研究开发支出 23.472 2.626 .610 8.939 .000 .218 4.581
a. Dependent Variable: 能源消耗量
Model Summary(b)
Model R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson
dimension0 1 .994a .989 .986 7713.63413 2.244
a. Predictors: (Constant), 研究开发支出, 第二产业比重, 第一产业比重
b. Dependent Variable: 能源消耗量
Eviews结果:
Dependent Variable: CA
Method: Least Squares
Date: 04/22/11 Time: 22:55
Sample: 1995 2009
Included observations: 15
Variable Coefficient Std. Error t-Statistic Prob.
C 1.45E+08 1.59E+08 0.913698 0.3824
I1 -1461060. 1586871. -0.920718 0.3789
I2 -1443129. 1590342. -0.907433 0.3855
I3 -1455130. 1588227. -0.916197 0.3811
RD 27.24001 4.889825 5.570753 0.0002
R-squared 0.989682 Mean dependent var 193661.5
Adjusted R-squared 0.985554 S.D. dependent var 64652.31
S.E. of regression 7770.561 Akaike info criterion 21.01527
Sum squared resid 6.04E+08 Schwarz criterion 21.25129
Log likelihood -152.6146 Hannan-Quinn criter. 21.01276
F-statistic 239.7877 Durbin-Watson stat 2.149388
Prob(F-statistic) 0.000000