How accurate is the political poll? When the First Lady Hillary Rodham Clinton and New York City Mayor Rudolph Giuliani. run for elective office. The race was highly publicized, historically noteworthy since it was the first time a First Lady has ever run for elective office. New Yorkers at that time had strong feelings about both candidates. Some polls showed Clinton having a slight lead, others showed Giuliani leading, but all were too close to call. How can polls taken around the same time, all using random sampling, suggest different conclusions about who is leading? How accurate are sample results and what is the "margin of error"? Please give your intuitive answers. Now the next phase is hands-on, collaborative learning activity. Let us simulate what happened. Let us become pollsters for a day, to simulate an actual poll that stated that, nationwide, 48% of American voters favored having Hillary Rodham Clinton run for the United States Senate. If 48% of the population wanted her to run, how likely would it be that a pollster would find that exactly 48% of a sample would say they wanted her to run? If a sample found a percentage other than 48%, how far off would the sample be likely to be? To explore these questions, you are welcome to make 100 tags, 48 of them marked "yes" and 52 of them marked "no." Repeated random sample ten tags and record the proportion of yeses in each sample, replacing each sample before drawing a new one. Generate 50 samples or just do as much as you can in twenty minutes. You may use a computer to simulate this process as well. Poll all the samples from the team members. Record your data. Show the frequency table (in relative frequency) of the outcomes.

Calculate the standard deviation of the proportions.

What is your explanation of what happened during the poll process?

Now repeat your experiment and draw 30 tags at a time instead of 10. What difference do you see? What have you learned about sampling distribution? Write a paper summarizing your findings.