The constant polls that come out during a presidential race are great examples of the ways that polls work. Essentially, they are what I like to call accurate estimates. In this election, the race in Ohio was incredibly tight. Polls have a sampling error and a confidence interval. That means that polls can simply determine that the election will be close. They have no way of determining who the winner will be though. There is simply too much overlap in the confidence intervals. For example, with a sampling error of +/- 3 three, there is a six percent confidence interval. Especially in a tight election, this is very large. The results of the poll could be very misleading so people need to keep the sampling error in mind. Without doubt though, the polls are still effective in gauging opinion. As this article shows, the polls conducted in Ohio were very accurate.