Bootstrap: how to obtain corrected coefficients?

[Svea Nolte]

My aim is to correct the coefficients that I obtained through quantile regression by using the following command:
. bsqreg BIOTHTOES_MIN DLEEFT LSLEN1, quantile(95) reps(200) BIOTHTOES_MIN is the dependent variable
DLEEFT and LSLEN1 are the independent variables My question is whether the coefficients in the result are already corrected by a certain 'shrinkage factor'?

[Maarten Buis]

You can correct for many things. So just saying you want to "correct" some coefficient does not uniquely identify what you want to do. If we don't know what you want to do, then it becomes very hard for us to help you.

We can answer your direct question: no, bsqreg only changes the standard errors. bsqreg gives you the same coefficients as qreg. This is something you could have easily confirmed yourself: just estimate the same model with qreg and compare the coefficients.

[Svea Nolte]

Sorry, that my question was too unclear:
I want to perform an internal validation of the quantile regression model to make an estimation of the performance of this model in future cohorts

Y = β₀ + β₁ * X + β₂ * Z

To validate the coefficients in this equation I performed a bootstrap.

As you suggested the same model with qreg showed the same coefficients.

Therefore, my question is how to obtain internal validated coefficients (corrected for too optimistic coefficients/ overfitting)?