用面板数据回归模型时，在固定效应输出窗口中进行F检验后，F检验的下面出现了新的固定效应的回归结果，和之前的不一样，到底以哪个为准呢？
Dependent Variable: LNSMOKE?   
Method: Pooled EGLS (Cross-section weights)   
Date: 09/26/10   Time: 17:30   
Sample (adjusted): 2001 2006   
Included observations: 6 after adjustments   
Cross-sections included: 30   
Total pool (balanced) observations: 180   
Iterate coefficients after one-step weighting matrix   
Convergence achieved after 12 total coef iterations   
   
Variable Coefficient Std. Error t-Statistic Prob.  
   
C 12.24410 0.020164 607.2283 0.0000
DLNEDU? -0.118406 0.047695 -2.482570 0.0142
DLNSCI?(-1) 0.078297 0.018017 4.345825 0.0000
DLNEDU?(-1) -0.243724 0.051277 -4.753119 0.0000
DLNEDU?(-2) -0.357171 0.077462 -4.610942 0.0000
DLNGOV?(-2) 0.232135 0.073315 3.166268 0.0019
AR(1) 0.466449 0.052718 8.847966 0.0000

上面的是初始的固定效应模型，下面是F检验的结果：

Redundant Fixed Effects Tests   
Pool: SMOKE   
Test cross-section fixed effects   
   
Effects Test                st              d.f.     prob
   
Cross-section F  43.651737 (29,144)  0. 00
接着有：
   
   
Cross-section fixed effects test equation:   
Dependent Variable: LNSMOKE?   
Method: Panel EGLS (Cross-section weights)   
Date: 09/26/10   Time: 17:58   
Sample (adjusted): 2001 2006   
Included observations: 6 after adjustments   
Cross-sections included: 30   
Total pool (balanced) observations: 180    s
se pre-specified GLS weights   
Convergence achieved after 9 total coef iterations   
   
Variable Coefficient Std. Error t-Statistic Prob.  
   
C 12.13240 0.150030 80.86625 0.0000
DLNEDU? -0.219731 0.078828 -2.787456 0.0059
DLNSCI?(-1) -0.007215 0.062036 -0.116298 0.9076
DLNEDU?(-1) -0.223520 0.093221 -2.397734 0.0176
DLNEDU?(-2) -0.873100 0.129039 -6.766197 0.0000
DLNGOV?(-2) 0.628100 0.141376 4.442751 0.0000
AR(1) 0.927017 0.025412 36.47961 0.0000
   
Weighted Statistics   
   
R-squared 0.885232     Mean dependent var  28.28588
Adjusted R-squared 0.881251     S.D. dependent var  26.47571
S.E. of regression 0.347575     Sum squared resid  20.89987
F-statistic 222.3978     Durbin-Watson stat  1.838411
Prob(F-statistic) 0.000000   
   
两部分红色字体部分不一致，pooled EGLS和panle EGLS有什么区别呢？由于两部分的参数检验结果不一致，影响最终结果，所以迫切想知道到底该以那个结果为准，望高手指教。