In: Math
In one exit poll of n = 140 voters, 66 said they voted for the
Democratic candidate and 74 said they voted for the Republican
candidate.
(a) Does a 95% confidence interval for the proportion voting for
the Democratic candidate allow you to predict the winner? Why or
why not?
No, because some of the values in the interval are negative (less than 0) or greater than 1, depending on whether we define the proportion to be voting for the Republican or Democratic candidate.No, because the interval includes a majority of people voting for the Democratic candidate and a majority of people voting for the Republican candidate. Yes, because the interval includes a majority of people voting for the Democratic candidate and a majority of people voting for the Republican candidate (proportions both above and below 0.5).Yes, because the interval doesn't include both values greater than 0.5 and values less than 0.5.No, because the interval doesn't include values greater than 0.5 (a majority of people voting for the Democratic candidate) and values less than 0.5 (a majority of people voting for the Republican candidate). Yes, because all the values in the interval are positive (greater than 0) and less than 1.
(b) A 95% confidence interval with n = 1400 voters and counts 660
and 740 would give different results than those above. Explain
why.
The larger sample size helps to reduce people's bias for one candidate or the other.The proportions of people who voted for the Democratic and Republican candidates would be different from those above. The z-scores in the confidence intervals would be different for this confidence interval from those above.We have a larger margin of error when we have a larger sample size, giving us more precision to estimate the parameter. The larger sample size provides more information, so when I have the same amount of confidence, I have more precision to estimate the parameter.