不太懂hausman 检验,所以不知道是我样本的问题还是什么问题,大家帮忙指点一二啊!
. tsset industry year
panel variable: industry, 1 to 5
time variable: year, 1996 to 2006
. xtreg tfp1 drd osrd ofrd, fe
Fixed-effects (within) regression Number of obs = 55
Group variable (i): industry Number of groups = 5
R-sq: within = 0.5045 Obs per group: min = 11
between = 0.9252 avg = 11.0
overall = 0.4198 max = 11
F(3,47) = 15.95
corr(u_i, Xb) = -0.8392 Prob > F = 0.0000
------------------------------------------------------------------------------
tfp1 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
drd | -.037854 .0225432 -1.68 0.100 -.0832051 .0074971
osrd | .0357088 .0257985 1.38 0.173 -.016191 .0876086
ofrd | -1.682817 .2831862 -5.94 0.000 -2.252514 -1.11312
_cons | 9.220949 .4751473 19.41 0.000 8.265076 10.17682
-------------+----------------------------------------------------------------
sigma_u | 10.349916
sigma_e | 1.8088003
rho | .9703625 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(4, 47) = 66.78 Prob > F = 0.0000
. est store fixed
. xtreg tfp1 drd osrd ofrd, re
Random-effects GLS regression Number of obs = 55
Group variable (i): industry Number of groups = 5
R-sq: within = 0.1572 Obs per group: min = 11
between = 0.9750 avg = 11.0
overall = 0.6148 max = 11
Random effects u_i ~ Gaussian Wald chi2(3) = 81.39
corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000
------------------------------------------------------------------------------
tfp1 | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
drd | .0956735 .0362738 2.64 0.008 .0245782 .1667688
osrd | -.1864348 .036152 -5.16 0.000 -.2572915 -.1155782
ofrd | 2.418761 .3145702 7.69 0.000 1.802215 3.035307
_cons | 9.550198 .9657279 9.89 0.000 7.657406 11.44299
-------------+----------------------------------------------------------------
sigma_u | 0
sigma_e | 1.8088003
rho | 0 (fraction of variance due to u_i)
------------------------------------------------------------------------------
. hausman fixed
---- Coefficients ----
| (b) (B) (b-B) sqrt(diag(V_b-V_B))
| fixed . Difference S.E.
-------------+----------------------------------------------------------------
drd | -.037854 .0956735 -.1335275 .
osrd | .0357088 -.1864348 .2221436 .
ofrd | -1.682817 2.418761 -4.101578 .
------------------------------------------------------------------------------
b = consistent under Ho and Ha; obtained from xtreg
B = inconsistent under Ha, efficient under Ho; obtained from xtreg
Test: Ho: difference in coefficients not systematic
chi2(3) = (b-B)'[(V_b-V_B)^(-1)](b-B)
= -2890.41 chi2<0 ==> model fitted on these
data fails to meet the asymptotic
assumptions of the Hausman test;
see suest for a generalized test
.
end of do-file