经管之家App
让优质教育人人可得
立即打开
做怀特检验，发现存在异方差，用加权最小二乘法以1/resid^2做权重修正，对修正后的方程做怀特检验，nR^2小于临界值通过了怀特检验，但是eviews输出结果中，变量中出现了WGT，不懂是什么意思呀，交叉项去哪了呀？这个结果是因为操作有误？我的模型是lny=C+C1LNX2+C2LNX4. 求大神解释~谢谢
| Heteroskedasticity Test: White |
| |||
|
|
|
|
|
|
|
|
|
|
|
|
| F-statistic | 0.509522 | Prob. F(2,23) | 0.6074 | |
| Obs*R-squared | 1.103089 | Prob. Chi-Square(2) | 0.5761 | |
| Scaled explained SS | 2.170983 | Prob. Chi-Square(2) | 0.3377 | |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| Test Equation: |
|
|
| |
| Dependent Variable: WGT_RESID^2 |
| |||
| Method: Least Squares |
|
| ||
| Date: 05/26/16 Time: 10:59 |
|
| ||
| Sample: 1989 2014 |
|
| ||
| Included observations: 26 |
|
| ||
| Collinear test regressors dropped from specification | ||||
|
|
|
|
|
|
|
|
|
|
|
|
| Variable | Coefficient | Std. Error | t-Statistic | Prob. |
|
|
|
|
|
|
|
|
|
|
|
|
| C | 6.21E-06 | 3.22E-06 | 1.928447 | 0.0662 |
| WGT^2 | 0.000355 | 0.000390 | 0.910210 | 0.3722 |
| LNX2^2*WGT^2 | -3.56E-06 | 3.92E-06 | -0.910239 | 0.3721 |
|
|
|
|
|
|
|
|
|
|
|
|
| R-squared | 0.042427 | Mean dependent var | 6.68E-06 | |
| Adjusted R-squared | -0.040841 | S.D. dependent var | 1.53E-05 | |
| S.E. of regression | 1.56E-05 | Akaike info criterion | -19.19343 | |
| Sum squared resid | 5.58E-09 | Schwarz criterion | -19.04826 | |
| Log likelihood | 252.5146 | Hannan-Quinn criter. | -19.15163 | |
| F-statistic | 0.509522 | Durbin-Watson stat | 2.123082 | |
| Prob(F-statistic) | 0.607406 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
胖胖小龟宝 发表于 2016-5-29 13:11
你的变量是WGT_RESID^2？如果是为了修正 你选择WLS就可以了啊Quick——Estimate Equation中的specification ...
扫码加好友，拉您进群
全部版块
我的主页
收藏
