How to extract robust standard errors in r?

Question

I calculated robust standard errors after running a regression with lm() function.

# robust standard errors
cov2I         <- vcovHC(ols2I, type = "HC1")
robust_se2I    <- sqrt(diag(cov2I))
print(robust_se2I)

I would like to extract the second value out of the resulting matrix and save it under a new variable. I've tried the following code but it didn't work.

stderrorols2I <- (summary(robust_se2I))[2]

Thanks for your help!

Could you please share your data using `dput()`? So we can reprdoce your problem. — Quinten, Apr 15 '22 at 13:07

score 2 · Accepted Answer · answered Apr 15 '22 at 13:10

This is how you could add RSE manually to your summary output. Alternativly, you could have a look at coeftest()

library(sandwich)  
mod1 <- lm(mpg ~ cyl + disp, data = mtcars) 

# robust standard errors
cov2I         <- vcovHC(mod1, type = "HC1")
robust_se2I    <- sqrt(diag(cov2I)) 

mod1 %>% 
  broom::tidy() %>% 
  mutate(rse = robust_se2I)

score 1 · Answer 2 · edited Apr 15 '22 at 13:42

1

I found the answer to my question.

To save the specific robust standard error I should have coded the following:

stderrorols2I <- (robust_se2I)[2]

Now it's working perfectly. Thanks anyway for your quick feedback!

edited Apr 15 '22 at 13:42

Nad Pat

3,129
3
10
20

answered Apr 15 '22 at 13:40

TFT

129
10

How to extract robust standard errors in r?

2 Answers2