During the recent election season, you saw many instances where polls were taken and reported, along with the confidence level of the estimate. For example, a poll showed that one candidate had a “”voting for” level of 37%+/-4. The opposing candidate had a score of 39%+/-4 at a 95% confidence level. Because the poll is not a census sample, a sample may be “good” but will have variation around the TRUE population proportion or mean. Thus, the poll is an estimate. The +/- value establishes a range within which the true population proportion is said to be, give a certain level of confidence.
However, the reporter said that the second candidate had “a slight lead”. Is this a true statement? Why could it be argued that the scores are “statistically the same”?
