Rasmussen took a special poll of one particular matchup possibility in the 2012 Presidential race. He looked at Ron Paul against Barack Obama and, surprise, Paul runs very well, trailing the incumbent President 41-42.
How can this be, when Paul was rejected so soundly in the Republican primaries just two years ago? Reading the poll, I'm left to wonder if it was written and paid for by supporters of the Texas Republican.
My suspicion, based the particulars of the poll that I find, is that the poll has a subtle bias toward Ron Paul. There is no indication that the questions are rotated or altered in any way, so I will assume they were asked as listed of all those surveyed.
First off, an opinion is only asked of Ron Paul, and not of Barack Obama, and it is asked first. This will influence further questions, including the Obama v. Paul question. A way to poll this that's more fair to the President would be to add a question asking for thoughts about Obama, and ask that one first half the time, while asking about Paul first the other half. This would negate the possible advantage Paul gets in the survey as asked.
Second, Ron Paul's name is listed first in the head to head question. This also adds bias. Given a list of names people are more likely to select an earlier name than a later name, an effect that has been seen even in major general elections, which is why states like California randomize the ballot order.
I don't even understand the point of the third question. The two choices given – that Ron Paul is either a divisive force or a new direction for the GOP – aren't even mutually exclusive. In fact, one might expect the two traits to correlate well. I haven't a clue how to interpret that question and its results in any meaningful way.
Lastly, we get to the fourth question, and the proposition is rejected that Ron Paul's values reflect those of the broader Republican Party by a 27-19 margin, MoE 3. It's 91% likely that the pool of likely voters agrees with this verdict, though of course the pool of likely voters also includes Democrats whose opinion of what Republican values are might differ from the opinion of Republican voters themselves. So again, I'm at a loss to understand the point of the question.
The entire poll seems driven by an agenda. Rasmussen doesn't say, but my suspicion is that the poll was commissioned and driven by activists in favor of Ron Paul, with the results skewed as a result.