if you go -robust- you have also taken into account possible autocorrelation of the epsilon residual.
That said, the most substantive post estimation test should be aimed at checking for possible model (or better regressand functional form) misspecification.
Take a look at the following toy-example, when an ancillary and an augmented regressions are run with -fitted- and -sq_fitted- values as predictors:
- [/size][/color][/font]
- [color=#252c2f][size=13px]webuse nlswork,clear[/size]
- [size=13px](National Longitudinal Survey. Young Women 14-26 years of age in 1968)[/size]
- [size=13px]
- [/size]
- [size=13px]. xtreg ln_wage age, fe robust[/size]
- [size=13px]
- [/size]
- [size=13px]Fixed-effects (within) regression Number of obs = 28,510[/size]
- [size=13px]Group variable: idcode Number of groups = 4,710[/size]
- [size=13px]
- [/size]
- [size=13px]R-sq: Obs per group:[/size]
- [size=13px] within = 0.1026 min = 1[/size]
- [size=13px] between = 0.0877 avg = 6.1[/size]
- [size=13px] overall = 0.0774 max = 15[/size]
- [size=13px]
- [/size]
- [size=13px] F(1,4709) = 884.05[/size]
- [size=13px]corr(u_i, Xb) = 0.0314 Prob > F = 0.0000[/size]
- [size=13px]
- [/size]
- [size=13px] (Std. Err. adjusted for 4,710 clusters in idcode)[/size]
- [size=13px]------------------------------------------------------------------------------[/size]
- [size=13px] | Robust[/size]
- [size=13px] ln_wage | Coef. Std. Err. t P>|t| [95% Conf. Interval][/size]
- [size=13px]-------------+----------------------------------------------------------------[/size]
- [size=13px] age | 0.018 0.001 29.73 0.000 0.017 0.019[/size]
- [size=13px] _cons | 1.148 0.018 64.81 0.000 1.113 1.183[/size]
- [size=13px]-------------+----------------------------------------------------------------[/size]
- [size=13px] sigma_u | .40635023[/size]
- [size=13px] sigma_e | .30349389[/size]
- [size=13px] rho | .64192015 (fraction of variance due to u_i)[/size]
- [size=13px]------------------------------------------------------------------------------[/size]
- [size=13px]
- [/size]
- [size=13px]. predict fitted, xb[/size]
- [size=13px](24 missing values generated)[/size]
- [size=13px]
- [/size]
- [size=13px]. g sq_fitted=fitted^2[/size]
- [size=13px](24 missing values generated)[/size]
- [size=13px]
- [/size]
- [size=13px]. xtreg ln_wage fitted sq_fitted , fe robust[/size]
- [size=13px]
- [/size]
- [size=13px]Fixed-effects (within) regression Number of obs = 28,510[/size]
- [size=13px]Group variable: idcode Number of groups = 4,710[/size]
- [size=13px]
- [/size]
- [size=13px]R-sq: Obs per group:[/size]
- [size=13px] within = 0.1087 min = 1[/size]
- [size=13px] between = 0.1006 avg = 6.1[/size]
- [size=13px] overall = 0.0865 max = 15[/size]
- [size=13px]
- [/size]
- [size=13px] F(2,4709) = 507.42[/size]
- [size=13px]corr(u_i, Xb) = 0.0440 Prob > F = 0.0000[/size]
- [size=13px]
- [/size]
- [size=13px] (Std. Err. adjusted for 4,710 clusters in idcode)[/size]
- [size=13px]------------------------------------------------------------------------------[/size]
- [size=13px] | Robust[/size]
- [size=13px] ln_wage | Coef. Std. Err. t P>|t| [95% Conf. Interval][/size]
- [size=13px]-------------+----------------------------------------------------------------[/size]
- [size=13px] fitted | 7.143 0.738 9.67 0.000 5.696 8.591[/size]
- [size=13px] sq_fitted | -1.816 0.219 -8.30 0.000 -2.245 -1.387[/size]
- [size=13px] _cons | -5.168 0.621 -8.32 0.000 -6.385 -3.950[/size]
- [size=13px]-------------+----------------------------------------------------------------[/size]
- [size=13px] sigma_u | .4039153[/size]
- [size=13px] sigma_e | .30245467[/size]
- [size=13px] rho | .64073314 (fraction of variance due to u_i)[/size]
- [size=13px]------------------------------------------------------------------------------[/size]
- [size=13px]
- [/size]
- [size=13px]. test sq_fitted[/size]
- [size=13px]
- [/size]
- [size=13px] ( 1) sq_fitted = 0[/size]
- [size=13px]
- [/size]
- [size=13px] F( 1, 4709) = 68.87[/size]
- [size=13px] Prob > F = 0.0000[/size]
- [size=13px]
- [/size]
- [size=13px]. xtreg ln_wage age fitted sq_fitted , fe robust[/size]
- [size=13px]note: fitted omitted because of collinearity[/size]
- [size=13px]
- [/size]
- [size=13px]Fixed-effects (within) regression Number of obs = 28,510[/size]
- [size=13px]Group variable: idcode Number of groups = 4,710[/size]
- [size=13px]
- [/size]
- [size=13px]R-sq: Obs per group:[/size]
- [size=13px] within = 0.1087 min = 1[/size]
- [size=13px] between = 0.1006 avg = 6.1[/size]
- [size=13px] overall = 0.0865 max = 15[/size]
- [size=13px]
- [/size]
- [size=13px] F(2,4709) = 507.42[/size]
- [size=13px]corr(u_i, Xb) = 0.0440 Prob > F = 0.0000[/size]
- [size=13px]
- [/size]
- [size=13px] (Std. Err. adjusted for 4,710 clusters in idcode)[/size]
- [size=13px]------------------------------------------------------------------------------[/size]
- [size=13px] | Robust[/size]
- [size=13px] ln_wage | Coef. Std. Err. t P>|t| [95% Conf. Interval][/size]
- [size=13px]-------------+----------------------------------------------------------------[/size]
- [size=13px] age | 0.130 0.013 9.67 0.000 0.103 0.156[/size]
- [size=13px] fitted | 0.000 (omitted)[/size]
- [size=13px] sq_fitted | -1.816 0.219 -8.30 0.000 -2.245 -1.387[/size]
- [size=13px] _cons | 3.034 0.229 13.23 0.000 2.585 3.484[/size]
- [size=13px]-------------+----------------------------------------------------------------[/size]
- [size=13px] sigma_u | .4039153[/size]
- [size=13px] sigma_e | .30245467[/size]
- [size=13px] rho | .64073314 (fraction of variance due to u_i)[/size]
- [size=13px]------------------------------------------------------------------------------[/size][/color]
- [color=#252c2f][size=13px]
As expected, the -test- outcome shows that the regression model is misspecified (because -age- taken as the unique predictor cannot give a fair and true view of the data generating process under investigation. Moreover, -age- has a non-linear relatinship with the regressand, as it can be easily found out by replacing -age- with -c.age##c.age-).
[code]
xtreg ln_wage c.age##c.age, fe robust
Fixed-effects (within) regression Number of obs = 28,510
Group variable: idcode Number of groups = 4,710
R-sq: Obs per group:
within = 0.1087 min = 1
between = 0.1006 avg = 6.1
overall = 0.0865 max = 15
F(2,4709) = 507.42
corr(u_i, Xb) = 0.0440 Prob > F = 0.0000
(Std. Err. adjusted for 4,710 clusters in idcode)
------------------------------------------------------------------------------
| Robust
ln_wage | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
age | 0.054 0.004 12.52 0.000 0.045 0.062
|
c.age#c.age | -0.001 0.000 -8.30 0.000 -0.001 -0.000
|
_cons | 0.640 0.062 10.25 0.000 0.518 0.762
-------------+----------------------------------------------------------------
sigma_u | .4039153
sigma_e | .30245467
rho | .64073314 (fraction of variance due to u_i)
------------------------------------------------------------------------------
[/code]