面板数据分析对决牛.ppt

0.000000 Prob(F-statistic) Unweighted Statistics 13.36561 Mean dependent var 0.986797 R-squared 0.146595 ? Durbin-Watson stat 1.505698 Sum squared resid Effects Specification S.D.?? Rho?? Cross-section random 0.123016 0.8009 Idiosyncratic random 0.061341 0.1991 Weighted Statistics R-squared 0.991988 ????Mean dependent var 1.706711 Adjusted R-squared 0.991709 ????S.D. dependent var 0.675234 S.E. of regression 0.061485 ????Sum squared resid 0.325115 F-statistic 3549.315 ????Durbin-Watson stat 0.678922 在表8.9的回归结果中，我们可以看到，横截面随机效应模型拟合得也很好，模型整体是显著的，回归系数的估计结果也是显著的，并且符号与预期一致。因此，我们到底应该选择固定效应模型还是随