Source | SS df MS Number of obs = 1827
-------------+------------------------------ F( 21, 1805) = 3.16
Model | .049369722 21 .002350939 Prob > F = 0.0000
Residual | 1.34474691 1805 .000745012 R-squared = 0.0354
-------------+------------------------------ Adj R-squared = 0.0242
Total | 1.39411663 1826 .000763481 Root MSE = .02729
------------------------------------------------------------------------------
bgscale_ | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
y | -.0029444 .0030846 -0.95 0.340 -.0089941 .0031053
cj | .0007467 .0008646 0.86 0.388 -.0009491 .0024425
marketindex | -.0000111 .0003287 -0.03 0.973 -.0006557 .0006335
TQ | -.0001398 .000203 -0.69 0.491 -.0005379 .0002583
independent | -.0022316 .0132557 -0.17 0.866 -.0282298 .0237666
gqjzd | .0153389 .0044251 3.47 0.001 .00666 .0240177
size | -.00063 .0008464 -0.74 0.457 -.0022899 .00103
year1 | -.0017837 .0019244 -0.93 0.354 -.0055579 .0019905
year2 | -.0026125 .0018541 -1.41 0.159 -.0062489 .0010239
year3 | -.0018585 .0017128 -1.09 0.278 -.0052178 .0015009
indcd_1 | -.0029144 .0046915 -0.62 0.535 -.0121158 .0062869
indcd_2 | -.0089599 .0115431 -0.78 0.438 -.0315992 .0136793
indcd_3 | -.0031042 .0027294 -1.14 0.256 -.0084574 .0022489
indcd_4 | .0216548 .0072966 2.97 0.003 .0073442 .0359654
indcd_5 | .0017719 .0051668 0.34 0.732 -.0083617 .0119055
indcd_6 | -.003426 .0061477 -0.56 0.577 -.0154834 .0086313
indcd_7 | .0035041 .0033793 1.04 0.300 -.0031235 .0101318
indcd_8 | -.0005758 .0038485 -0.15 0.881 -.0081237 .0069721
indcd_9 | .0115661 .0036584 3.16 0.002 .0043909 .0187413
indcd_10 | -.0054573 .0051087 -1.07 0.286 -.0154769 .0045623
indcd_11 | .01048 .0160415 0.65 0.514 -.0209819 .0419418
_cons | .0177281 .0203724 0.87 0.384 -.0222279 .0576842
------------------------------------------------------------------------------
. estat hettest
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
Ho: Constant variance
Variables: fitted values of bgscale_
chi2(1) = 384.49
Prob > chi2 = 0.0000
这个结果是不是表示有异方差问题呢,我后面分别用了WLS加权最小二乘法和white的方法(后面加robust),回归结果差异很大,wls效果更好,但white的和原回归差异不大,是怎么回事呢?