### Bootstrap: p
###
### Consider a poll related the the
Presidential Election of 790 voters. The results
### give 411 to Bush and 379 to Gore.
Thus 0.52 favor Bush and 0.48 favor Gore.
### Get a 95% CI for the proportion of
voters who favor Bush using the traditional
### CLT formula, p.hat +/- z * sqrt(p.hat*(1-p.hat)/n),
where z = 1.96.
### Using Minitab, Stat > Basic
Statistics > 1 Proportion > Summarized data > Number
### of trials: 790 > Number of
successes: 411
### Use the Bootstrap to plot the
Bootstrap distibution of p.hat and give a 95% Boostrap
### Empirical Confidence Interval for p.
# create a data vector
n <- 790
b <- 411
g <- n - b
x <- numeric(n)
x[1:b] <- 1
mean(x)
boot.b <- bootstrap(x,mean)
summary(boot.b)
limits.emp(boot.b)
plot(boot.b)