发布时间： 来源：人大经济论坛

最近在做面板数据分析，关于随机与固定的选择，我做了Hausman Test，总体可能只有0.3626，是不是表示我可以按固定效应面板模型来做了？请好心人回答，论文急着用，谢谢了！
Correlated Random Effects - Hausman Test
Pool: POOL
Test cross-section random effects

Test SummaryChi-Sq. Statistic Chi-Sq. d.f. Prob.

Cross-section random4.334302 4 0.3626

** Warning: robust standard errors may not be consistent with
assumptions of Hausman test variance calculation.

Cross-section random effects test comparisons:

Variable Fixed RandomVar(Diff.)Prob.

BK? 0.365886 0.352976 0.003209 0.8197
BL? 1.056029 0.944456 0.024025 0.4716
BE? -27.685875 -19.833824 30.130816 0.1526
BC? 28.833922 20.677856 32.309096 0.1513


Cross-section random effects test equation:
Dependent Variable: C?
Method: Panel Least Squares
Date: 11/12/09 Time: 16:58
Sample: 1997 2007
Included observations: 11
Cross-sections included: 7
Total pool (balanced) observations: 77
Cross-section SUR (PCSE) standard errors & covariance (d.f. corrected)

Coefficient Std. Error t-Statistic Prob.

C -0.028539 0.041353 -0.690125 0.4925
BK? 0.365886 0.104872 3.488887 0.0009
BL? 1.056029 0.307420 3.435136 0.0010
BE? -27.68588 10.24214 -2.703133 0.0087
BC? 28.83392 10.60113 2.719891 0.0083

Effects Specification

Cross-section fixed (dummy variables)

R-squared 0.471651 Mean dependent var0.127289
Adjusted R-squared 0.391598 S.D. dependent var0.071359
S.E. of regression 0.055661 Akaike info criterion-2.807528
Sum squared resid 0.204474 Schwarz criterion-2.472699
Log likelihood 119.0898 Hannan-Quinn criter.-2.673599
F-statistic 5.891735 Durbin-Watson stat1.693143
Prob(F-statistic) 0.000003