本人做的是panel data, 15年的数据, 5个自变量。对hausman检验结果有很多疑问,诚心向前辈求教!
1.第一个表,这样的实验结果代表豪斯曼检验无效吗?是不是可以理解选择固定效应模型?
Test Summary Chi-Sq. Statistic Chi-Sq. d.f. Prob.
Cross-section random 0.000000 6 1.0000
* Cross-section test variance is invalid. Hausman statistic set to zero.
2.这张表的Var(diff)和prob是什么意思?
Cross-section random effects test comparisons:
Variable Fixed Random Var(Diff.) Prob.
?UINCGR 0.120639 0.147496 0.000022 0.0000
?YDR -0.458071 -0.377785 0.000214 0.0000
?ADR 0.481880 0.341505 0.002898 0.0091
?GDPGR 0.010477 0.038480 0.000033 0.0000
?I -0.028842 -0.188186 0.000689 0.0000
?SMR -0.021760 0.079505 0.000379 0.0000
3.最后这张表又需要看哪些重要数据,它的作用是什么?
Cross-section random effects test equation:
Dependent Variable: ?USR
Method: Panel Least Squares
Sample: 1996 2010
Included observations: 15
Cross-sections included: 30
Total pool (balanced) observations: 450
Coefficient Std. Error t-Statistic Prob.
C 29.93427 2.032243 14.72967 0.0000
?UINCGR 0.120639 0.043786 2.755175 0.0061
?YDR -0.458071 0.036793 -12.44996 0.0000
?ADR 0.481880 0.126746 3.801946 0.0002
?GDPGR 0.010477 0.025963 0.403555 0.6867
?I -0.028842 0.089099 -0.323712 0.7463
?SMR -0.021760 0.289182 -0.075245 0.9401
Effects Specification
Cross-section fixed (dummy variables)
R-squared 0.707699 Mean dependent var 23.16849
Adjusted R-squared 0.682988 S.D. dependent var 5.321455
S.E. of regression 2.996183 Akaike info criterion 5.109173
Sum squared resid 3716.524 Schwarz criterion 5.437913
Log likelihood -1113.564 Hannan-Quinn criter. 5.238742
F-statistic 28.63857 Durbin-Watson stat 0.655268
Prob(F-statistic) 0.000000