eviews6.0
如何对面板数据进行上述检验呢，我在view里没有找到residuals test
DW也不符合
另，选择了固定效应模型，但估计结果和相关分析相差很多，比如明显的负相关变成了正相关，不知道是不是应该在什么检验后修正
以下是结果作为参考
Method: Pooled EGLS (Cross-section weights)   
Date: 07/18/10   Time: 22:43   
Sample: 2007 2009   
Included observations: 3   
Cross-sections included: 53   
Total pool (balanced) observations: 159   
Linear estimation after one-step weighting matrix   
White cross-section standard errors & covariance (d.f. corrected)   
WARNING: estimated coefficient covariance matrix is of reduced rank   
   
Variable Coefficient Std. Error t-Statistic Prob.  
   
C -10.70111 3.648565 -2.932963 0.0042
TOP? 0.006327 0.020407 0.310047 0.7572
TEN? 0.007002 0.014290 0.490008 0.6253
Z? -0.010596 0.000484 -21.89033 0.0000
SLS? 5.356698 2.722944 1.967245 0.0521
LNSAL? 1.122808 0.490039 2.291261 0.0242
MO? 0.370236 0.063151 5.862689 0.0000
LNTA? 0.648124 0.303260 2.137187 0.0352
ALR? 0.018895 0.007365 2.565488 0.0119
SCALE? -0.189526 0.053380 -3.550467 0.0006
ID? -21.35497 1.913728 -11.15883 0.0000
CEO? 1.618041 0.768205 2.106261 0.0378
Effects Specification   
   
Cross-section fixed (dummy variables)   
   
Weighted Statistics   
   
R-squared 0.951425     Mean dependent var  16.22863
Adjusted R-squared 0.919213     S.D. dependent var  19.65771
S.E. of regression 2.664376     Sum squared resid  674.3956
F-statistic 29.53579     Durbin-Watson stat  2.706655
Prob(F-statistic) 0.000000   
   
Unweighted Statistics   
   
R-squared 0.816940     Mean dependent var  8.035283
Sum squared resid 738.8596     Durbin-Watson stat  2.748124