頑健な推定方法(アウトプット）
Source | SS df MS Number of obs = 30
-------------+------------------------------ F( 1, 28) = 2698.46
Model | 100148.521 1 100148.521 Prob > F = 0.0000
Residual | 1039.17062 28 37.1132363 R-squared = 0.9897
-------------+------------------------------ Adj R-squared = 0.9894
Total | 101187.692 29 3489.23074 Root MSE = 6.0921
------------------------------------------------------------------------------
cp | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
yd | .8957558 .0172438 51.95 0.000 .8604336 .931078
_cons | -11.42457 3.66786 -3.11 0.004 -18.93784 -3.911295
------------------------------------------------------------------------------
. dwstat;
Durbin-Watson d-statistic( 2, 30) = .1703267
. /* Durbin Watson Test */
> hettest;
Cook-Weisberg test for heteroskedasticity using fitted values of cp
Ho: Constant variance
chi2(1) = 0.93
Prob > chi2 = 0.3339
. /* Cook-Weisberg test for heteroskedasticity */
> ovtest;
Ramsey RESET test using powers of the fitted values of cp
Ho: model has no omitted variables
F(3, 25) = 51.55
Prob > F = 0.0000
. /* Ramsey RESET 4 */
> prais cp yd;
Iteration 0: rho = 0.0000
Iteration 1: rho = 0.8888
Iteration 2: rho = 0.9404
Iteration 3: rho = 0.9538
Iteration 4: rho = 0.9572
Iteration 5: rho = 0.9581
Iteration 6: rho = 0.9583
Iteration 7: rho = 0.9583
Iteration 8: rho = 0.9583
Iteration 9: rho = 0.9583
Iteration 10: rho = 0.9583
Prais-Winsten AR(1) regression -- iterated estimates
Source | SS df MS Number of obs = 30
-------------+------------------------------ F( 1, 28) = 31.36
Model | 182.633882 1 182.633882 Prob > F = 0.0000
Residual | 163.042912 28 5.82296116 R-squared = 0.5283
-------------+------------------------------ Adj R-squared = 0.5115
Total | 345.676795 29 11.9198895 Root MSE = 2.4131
------------------------------------------------------------------------------
cp | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
yd | .8404458 .0427162 19.68 0.000 .7529455 .927946
_cons | 2.705897 10.72644 0.25 0.803 -19.26621 24.678
-------------+----------------------------------------------------------------
rho | .9583424
------------------------------------------------------------------------------
Durbin-Watson statistic (original) 0.170327
Durbin-Watson statistic (transformed) 1.194382
. /* Iterative GLS */
> regress cp yd, robust;
Regression with robust standard errors Number of obs = 30
F( 1, 28) = 2972.07
Prob > F = 0.0000
R-squared = 0.9897
Root MSE = 6.0921
------------------------------------------------------------------------------
| Robust
cp | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
yd | .8957558 .0164308 54.52 0.000 .8620988 .9294129
_cons | -11.42457 3.777366 -3.02 0.005 -19.16215 -3.686983
------------------------------------------------------------------------------
. /* White Robust SE */
> newey cp yd, lag(0) ;
Regression with Newey-West standard errors Number of obs = 30
maximum lag : 0 F( 1, 28) = 2972.07
Prob > F = 0.0000
------------------------------------------------------------------------------
| Newey-West
cp | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
yd | .8957558 .0164308 54.52 0.000 .8620988 .9294129
_cons | -11.42457 3.777366 -3.02 0.005 -19.16215 -3.686983
------------------------------------------------------------------------------
. /* No Autocorrelation in Error */
> newey cp yd, lag(2);
Regression with Newey-West standard errors Number of obs = 30
maximum lag : 2 F( 1, 28) = 1199.36
Prob > F = 0.0000
------------------------------------------------------------------------------
| Newey-West
cp | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
yd | .8957558 .0258651 34.63 0.000 .8427735 .9487381
_cons | -11.42457 5.941265 -1.92 0.065 -23.5947 .7455642
------------------------------------------------------------------------------
. /* 2nd order autocorrelation in Error */