Home > Archives > Election Insider
Two Pollsters, Two Different Results
September 9, 2010 Content provided by DDC Advocacy
The new CNN/Time Magazine (9/2-7; 869 registered voters) and Rasmussen Reports (9/7; 500 likely voters) surveys of the Kentucky Senate race were both released yesterday, but the two polls come to very different conclusions. Where Rasmussen is giving Republican Rand Paul his largest lead of the campaign, 51-38% over Attorney General Jack Conway - 54-39% when respondents who say they are leaning to a particular candidate are added - CNN/Time shows the race to be a dead heat. According to the latter results, both candidates are tied at 46%. Why the difference? Though Rasmussen has the smaller sampling pool (500 to CNN/Time's 869), and usually having a large universe is more desirable, their poll is taken over a concentrated period of time: one day as compared to six. Most polls are usually conducted during a three-day window, so doubling that time factor, as in CNN/Time's study, does hurt the sample's reliability. Rasmussen also utilizes a formula that determines which members of the original universe are likely to vote, and then develops the actual polling sample from the larger pool. The CNN/Time data is only based upon surveying registered voters, regardless of vote propensity. We'll have to wait until Election Day to find out exactly who is right but, today, the Rasmussen data is probably more accurate.