UC知道请问：做计量经济学论文，数据是否可能不存在异方差性？
来源：百度知道 编辑：UC知道 时间：2024/05/06 02:41:40
用EVIEWS进行怀特检验，得出数据：
White Heteroskedasticity Test:
F-statistic 2.425769 Probability 0.087784
Obs*R-squared 9.283873 Probability 0.098263
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 06/26/08 Time: 13:47
Sample: 1987 2006
Included observations: 20
Variable Coefficient Std.Error t-Statistic Prob.
C 1298242. 1460166. 0.889106 0.3890
X1 -6743.641 6876.046 -0.980744 0.3434
X1^2 8.112141 6.218027 1.304617 0.2131
X1*X2 -0.061513 0.070592 -0.871388 0.3982
X2 33.40899 62.25825 0.536619 0.6000
X2^2 6.69E-05 3.73E-05 1.795267 0.0942
R-squared 0.464194 Mean dependent var 241824.4
Adjusted R-squared 0.272834 S.D. dependent var 306557.5
S.E. of regression 261413.9 Akaike info criterion28.02892
Sum squa
White Heteroskedasticity Test:
F-statistic 2.425769 Probability 0.087784
Obs*R-squared 9.283873 Probability 0.098263
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 06/26/08 Time: 13:47
Sample: 1987 2006
Included observations: 20
Variable Coefficient Std.Error t-Statistic Prob.
C 1298242. 1460166. 0.889106 0.3890
X1 -6743.641 6876.046 -0.980744 0.3434
X1^2 8.112141 6.218027 1.304617 0.2131
X1*X2 -0.061513 0.070592 -0.871388 0.3982
X2 33.40899 62.25825 0.536619 0.6000
X2^2 6.69E-05 3.73E-05 1.795267 0.0942
R-squared 0.464194 Mean dependent var 241824.4
Adjusted R-squared 0.272834 S.D. dependent var 306557.5
S.E. of regression 261413.9 Akaike info criterion28.02892
Sum squa
你好，你的P是0.087784，即拒绝原假设犯错的概率是8.78%。怀特检验的原假设是假设不存在异方差，那么P=0.087784的意思就是：
承认原模型存在异方差，犯错的概率是8.78%。因此，如果你想跳过加权过程直接进行序列相关检验，一定要说名前提，即“在1%（或5%）的置信区间内不存在异方差”。置信区间取至10%的时候就仍要做加权。
小弟太佩服两位了。