Selecting the best model with information criterion

[Iva Mihaylova]

Hello,

I am running models of the type

foreach outcome of varlist a b c {
foreach tree of varlist eth_frc_t1-eth_frc_t13 {
eststo TR1_2_`y' : reg `outcome' `tree' $main_controls , ro
esttab TR1* using Result_1.csv, append b(3) se(3) ar2(3) ///
se star(^{*} 0.1 ^{**} 0.05 ^{***} 0.01) title(Title)
eststo clear
}
}

My independent variable has 13 possible levels of compilation and, therefore, is eth_frc_t1-eth_frc_t13.

Is there any proper way of selecting the best fitting model through AIC or other model fit parameters? How would this be calculated in STATA?

[Sebastian Kripfganz]

After the regress command, use the postestimation command estat ic for the model information criteria.

[Iva Mihaylova]

After the regress command, use the postestimation command estat ic for the model information criteria.

Thank you!