Pfthb

For a stats course I am teaching (in both R and Stata) I want to replicate Stata's robust command in R. This can be done easily using the coeftest command from the lmtest package. Using an example with data from the car library, I get the following

library(car)
library(lmtest)
library(sandwich)
mod <- lm(prestige ~ education + income, data=Prestige)
coeftest(mod, vcov = vcovHC(mod, "HC1"))

This replicates what Stata calls robust regression. But now I want to use the effects package and obtain predicted values (and confidence intervals) for this regression using the heteroscedasticity robust standard errors. According to the documentation of the package, I should be able to use the vcov. argument to change the covariance matrix to the robust one. However, when I try this as follows

effect("education", mod, vcov.=vcovHC(mod, "HC1"))

I get the error message

Error in vcov.(mod) : could not find function "vcov."

I also tried the hccm command instead, as this is from the same author as the effects package, but I get the same error message. I can't figure out where my mistake is. Any thoughts?