1
$\begingroup$
set.seed(1234)
x <- c(rnorm(100, mean = 50, sd=150))
group <- rep(0:1, each=50)

df <- data.frame(group, x)

df$group <- as.factor(df$group)

t.test(df$x ~ df$group, paired=T)

boot.diff <- function(d, i) {
  d$group <- d$group[i]
  mean = tapply(d$x, INDEX = d$group, mean)
  Diff = mean[1] - mean[2]
  return(Diff)
}

boot(df, boot.diff, R=1000)
boot.ci(boot(df, boot.diff, R=1000))
plot(boot(df, boot.diff, R=1000))

Attached above is my code - I'm fairly new to the idea of bootstrapping statistics - In the above code I'm trying to get the bootstrapped difference between two means, similar to a paired t-test.

Can someone please explain the various differences in the confidence intervals produced from the boot package, i.e "normal" "percentile" "Bca" "Basic"

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.