How Stupid Does Rassmussen Think We Are?

Sister Toldjah gets a little down about a Rasmussen poll that shows Obama in a statistical tie with McCain in North Carolina. So I went and took a look at the survey to see if she really should be that down. Well, she shouldn't. Here are the first two questions:

1* How do you rate the way that George W. Bush is performing his role as President? Excellent, good, fair, or poor?

2*If the Presidential Election were held today, would you vote for Republican John McCain or Democrat Barack Obama?

Anything strike you as odd? Well, yeah, me too. What the hell is George Bush doing at the top of the survey? I didn't know he was running for anything! Hmmm...that's queer.

Well, actually it isn't. It's called "priming." That is when in polling you ask a question which puts people into a state of mind that influences the way they answer follow-up questions. Obviously, Rasmussen is gonna claim that the George Bush question was merely there to be used in the crosstabs, i.e. various demographic breakdowns of the survey sample responses. But, if you were going to do that why not have the Bush question after the "Were the election held today" question?

It's simple. If you were just interested in getting a snapshot of state's opinion, you would have asked the Bush question later. However, if you wanted to prime your audience, by reminding them of an unpopular President, AND reminding voters in the second question just which candidate is also a Republican, well then you'd put the Bush question first.

Look, these are real things that reputable Political Scientists always keep an eye out for. By and large, they are actually interested in generating the best data possible, so they make sure that lousy question wording or leading question ordering don't contaminate their results. That's what honest people do, because sometimes a survey can get messed up accidently if you don't.

Rasmussen, on the other hand, seem to be unconcerned about such things.