In the year leading up to a big election, there are many polls conducted to estimate what percent of the population is likely to vote for each candidate. The true percent can change during the year, so the polls are not necessarily estimating the same thing every time. Suppose 100 such polls are done and each one is used to find a 95% confidence interval for the percent of the population who plan to vote for a certain candidate at the time of the poll.
a. About how many of the 100 confidence intervals are likely to cover the true percent?
b. Is there any way to know which of the polls cover the true percent and which ones don't?