Use StatKey or other technology to create a randomization distribution for the poll described in Exercise 4.137. What is the p-value for the test? Would this have been convincing evidence that Obama would win the election (at least the popular vote)? Now, knowing the true outcome of the election, does the test avoid making a Type I or Type II error?

Exercise 4.137

In October before the 2008 US presidential election, ABC News and the Washington Post jointly conducted a poll of ‘‘a random national sample” and asked people who they intended to vote for in the 2008 presidential election. Of the 1057 sampled people who answered either Barack Obama or John McCain, 55.2% indicated that they would vote for Obama while 44.8% indicated that they would vote for McCain. While we now know the outcome of the election, at the time of the poll many people were very curious as to whether this significantly predicts a winner for the election. (While a candidate needs a majority of the electoral college vote to win an election, we’ll simplify things and simply test whether the percentage of the popular vote for Obama is greater than 50%.)