蓝色 发表于 2007-2-8 08:17 
1、F检验,可以检验到底是pooled ols还是fixed model
2、xttest0 是检验到底是pooled ols还是random model ...
蓝色,老师,您好:下面是我做的2个不同回归:
第一个
xtivreg2 lnxf you old lninf ( lnsr= lnf ) ,fe
FIXED EFFECTS ESTIMATION
------------------------
Number of groups = 30 Obs per group: min = 8
avg = 8.0
max = 8
IV (2SLS) estimation
--------------------
Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only
Number of obs = 240
F( 4, 206) = 0.45
Prob > F = 0.7712
Total (centered) SS = .5746370205 Centered R2 = -12.5247
Total (uncentered) SS = .5746370205 Uncentered R2 = -12.5247
Residual SS = 7.771810297 Root MSE = .1924
------------------------------------------------------------------------------
lnxf | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
lnsr | -5.844567 43.90304 -0.13 0.894 -91.89295 80.20381
you | .009729 .0480343 0.20 0.839 -.0844165 .1038745
old | .0043544 .0247127 0.18 0.860 -.0440817 .0527905
lninf | .1397295 .6115373 0.23 0.819 -1.058862 1.338321
------------------------------------------------------------------------------
Underidentification test (Anderson canon. corr. LM statistic): 0.021
Chi-sq(1) P-val = 0.8848
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic): 0.021
Stock-Yogo weak ID test critical values: 10% maximal IV size 16.38
15% maximal IV size 8.96
20% maximal IV size 6.66
25% maximal IV size 5.53
Source: Stock-Yogo (2005). Reproduced by permission.
------------------------------------------------------------------------------
Sargan statistic (overidentification test of all instruments): 0.000
(equation exactly identified)
------------------------------------------------------------------------------
Instrumented: lnsr
Included instruments: you old lninf
Excluded instruments: lnf
------------------------------------------------------------------------------
第二个
. xtivreg2 lnxf you old lninf ( lnsr= lnf ) ,fe r
FIXED EFFECTS ESTIMATION
------------------------
Number of groups = 30 Obs per group: min = 8
avg = 8.0
max = 8
IV (2SLS) estimation
--------------------
Estimates efficient for homoskedasticity only
Statistics robust to heteroskedasticity
Number of obs = 240
F( 4, 206) = 0.58
Prob > F = 0.6789
Total (centered) SS = .5746370205 Centered R2 = -12.5247
Total (uncentered) SS = .5746370205 Uncentered R2 = -12.5247
Residual SS = 7.771810297 Root MSE = .1924
------------------------------------------------------------------------------
| Robust
lnxf | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
lnsr | -5.844567 38.34995 -0.15 0.879 -81.00909 69.31996
you | .009729 .0411983 0.24 0.813 -.0710182 .0904762
old | .0043544 .0195486 0.22 0.824 -.03396 .0426689
lninf | .1397295 .5189498 0.27 0.788 -.8773934 1.156852
------------------------------------------------------------------------------
Underidentification test (Kleibergen-Paap rk LM statistic): 0.026
Chi-sq(1) P-val = 0.8722
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic): 0.021
(Kleibergen-Paap rk Wald F statistic): 0.025
Stock-Yogo weak ID test critical values: 10% maximal IV size 16.38
15% maximal IV size 8.96
20% maximal IV size 6.66
25% maximal IV size 5.53
Source: Stock-Yogo (2005). Reproduced by permission.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments): 0.000
(equation exactly identified)
------------------------------------------------------------------------------
Instrumented: lnsr
Included instruments: you old lninf
Excluded instruments: lnf
不明白,第一个回归中的识别不足检验
Anderson canon. corr. LM statistic与第二个回归中的Kleibergen-Paap rk LM statistic有什么区别?同样是
(Cragg-Donald Wald F statistic 和 Kleibergen-Paap rk Wald F statistic的区别?以及
Sargan statistic 与 Hansen J statistic 的区别是什么?