A research study conducted by Copernicus Marketing Consulting and Research back in 2008 about the phenomenon of polling failure and response bias related to the presidential election. The link is here:
I found this to relate to what we have been learning in class because they talk about the issue known to be a common mistake in research studies: response bias. More so, the study was interesting to me personally because it was related to the presidential election and how the polls are not always correct. I find this topic interesting to learn about, so I wanted to see what they discovered in their study.
The article mentions that it is historic that the polls leading up the the election typically are not correct, or there is no pattern of them being so. Polls are not usually accurate because the people who conduct them have typically had a following more dominated by either Democratic or Republican voters. This makes their results swayed one way or the other because of response bias; the people who are responding to the poll is not a collective sample of the entire U.S. population.
It is interesting to me, and they even say it in the article, why pollsters disregard this response bias in their polls. They continue to conduct polls the same and do not seem to care about the bias that shows through and the unreliable results they find for each election. I would think there would be interest in yielding an accurate poll, but I suppose that would be extremely difficult to do for the entire United States population.