With the Presidential Campaign Season in full swing, there is something that you should keep in mind as you look at the daily Presidential Poll results and that’s to check the polls sampling numbers. The numbers polled from each party. This past week there have been Presidential Polls released that have been over sampling the Democrat Party. Obviously Democrats are more than likely going to vote Obama so he will look as though he has the momentum. Pollers say they over sample to take into account voter motivation (which party is more likely to turnout the most voters). This begs the question, will Democrats turnout in huge numbers when their party and their president really haven’t done anything to turn the economy around? Isn’t it possible they might be somewhat melancholy? This time around I believe Republicans (Conservatives) are much more motivated to remove what they perceive as a very inept and socialistic president. To buttress my point: “…Gallup just released a survey that shows Democrat voter enthusiasm is at its lowest point in years and Republican enthusiasm is a full 12 points higher.” Gallup is generally considered somewhat of a left leaning polling organization.
The most recent example of skewed poll results is the latest PEW Research Center Poll. They give Obama and 51% to 41% lead over Romney. I’m not sure if they are incompetent, paid to skew results or just plain partisan but they sure look stupid. The sampling numbers: Republicans 459 Democrats 813 and Independents 599. They sampled 354 more Democrats than Republicans. To better understand why you need to look at the sampling read this article.
Earlier in the week the New York Times – Quinnipiac Poll skewed the samples as well. Their poll focused on three “Swing States,” Florida, Ohio and Pennsylvania. More about that here. Allahpundit at HotAir.Com has some thoughts about skewing in these two polls as well.
Other things to keep in mind when looking at polls include who is sponsoring (paying) for the poll, how the questions are asked, and are the voters likely voters or registered voters? Likely voters may provide a more accurate measurement because the voters are “more likely to vote” than someone who is registered. Gallup explains some about their process here.
I like these two polling organizations:
The next time you read a poll or hear a newsperson or someone you know rattle off the stats…..Think about the Details.