vcovHC and confidence interval

Question

Is it possible to have confint use the robust vcov obtained by vcovHC (from the sandwich package) after fitting a model?

Answer 1

Nope, you cannot use the function confint directly with the robust vcov. But it's pretty straight-forward to do this by hand.

x <- sin(1:100)
y <- 1 + x + rnorm(100)
## model fit and HC3 covariance
fm <- lm(y ~ x)
Cov <- vcovHC(fm)

tt <-qt(c(0.025,0.975),summary(fm)$df[2])
se <- sqrt(diag(Cov))
ci <-coef(fm) + se %o% tt

Otherwise, your can adapt the confint.default() function to your own needs :

confint.robust <- function (object, parm, level = 0.95, ...)
{
    cf <- coef(object)
    pnames <- names(cf)
    if (missing(parm))
        parm <- pnames
    else if (is.numeric(parm))
        parm <- pnames[parm]
    a <- (1 - level)/2
    a <- c(a, 1 - a)
    pct <- stats:::format.perc(a, 3)
    fac <- qnorm(a)
    ci <- array(NA, dim = c(length(parm), 2L), dimnames = list(parm,
        pct))
    ses <- sqrt(diag(sandwich::vcovHC(object)))[parm]
    ci[] <- cf[parm] + ses %o% fac
    ci
}

As Brandon already suggested, you'd get more chance of a quick answer if you ask these things at stats.stackexchange.com