White Heteroskedasticity Test: |
||||
F-statistic |
4.969152 |
Probability |
0.014873 |
|
Obs*R-squared |
8.019599 |
Probability |
0.018137 |
|
Test Equation: |
||||
Dependent Variable: RESID^2 |
||||
Method: Least Squares |
||||
Date: 05/01/07 Time: 23:54 |
||||
Sample: 1 29 |
||||
Included observations: 29 |
||||
Variable |
Coefficient |
Std. Error |
t-Statistic |
Prob. |
C |
-219.6978 |
401.7777 |
-0.546814 |
0.5892 |
X |
0.159502 |
0.107465 |
1.484227 |
0.1498 |
X^2 |
-3.54E-06 |
5.95E-06 |
-0.595011 |
0.5570 |
R-squared |
0.276538 |
Mean dependent var |
681.7481 |
|
Adjusted R-squared |
0.220887 |
S.D. dependent var |
985.7066 |
|
S.E. of regression |
870.0574 |
Akaike info criterion |
16.47269 |
|
Sum squared resid |
19681997 |
Schwarz criterion |
16.61414 |
|
Log likelihood |
-235.8540 |
F-statistic |
4.969152 |
|
Durbin-Watson stat |
2.074317 |
Prob(F-statistic) |
0.014873 |
|
请问这样的结果怎样用p-值判断是否具有异方差,如果显著性水平是0.01的话应该是什么样子?0.05呢?
这与OLS估计量的t检验判断规则正好相反吗 ?为什么?
nr^2>F值,故存在异方差.
或者不确定的话还可以用PARK &GLEJSER检验,或者B-P检验来验证是否存在异方差.
转载请注明:数据分析 » 关于Eviews异方差White检验_eviews异方差检验