I have a `quantile regression model`

that I fit with the `rq()`

function in the `quantreg`

package in R.

However, since my sample size if fairly small (n = 36) compared to the number of X variables (= 8), I need to estimate power for the various regression coefficients. How can I do that?

It would be great to be pointed to a function in R that can do that for a `quantreg model`

, but a general explanation of how to calculate power in such a model would be fine as well (I can then code that in R myself).

How can I determine the power for a given `quantile regression model`

with a given dataset?

**Contents**hide

#### Best Answer

Sorry, I don't know what you say clearly. If you want to compare different rq model with different variable that which is better, AIC and BIC are common ways, no matter what your regression model is.

In R:

`fit1 <- rq(...) fit2 <- rq(...) AIC(fit1) AIC(fit2) BIC(fit1) BIC(fit2) ## briefly AIC(fit1, fit2) BIC(fit1, fit2) `

And if you want to check which variable is more important to effect the results(feature selection), I have no idea about so.

### Similar Posts:

- Solved – (Quantile regression) Which standard error for heteroscedasticity & serial correlation
- Solved – Goodness of fit tests for quantile regression in R
- Solved – Goodness of fit tests for quantile regression in R
- Solved – ARIMA quantile regression in R
- Solved – Quantile regression for Sales forecast in R