Rasmussen is known for asking issue questions in a manner which helps get the response he wants (i.e., he often frames the question in a way which is beneficial to the answer he prefers, like by asking a question “some people argue that X will do Y. do you agree?” instead of just asking “do you support X”).
You see this in his latest poll asking about racism in the campaign as well.
What are the questions about the Paris/Brittney/Obama ad?
“Have you read, seen, or heard news reports about the television commercial including images of Britney Spears and Paris Hilton?”
“Have you actually seen the commercial?”
“Is the McCain ad racist?”
OK, generally fair enough…if he followed the same trend when asking about Obama, but he didn’t. Here are the question he asked about Obama’s comment:
“Barack Obama said his opponent will try to scare people because Obama does not look like the other Presidents on dollar bills. Is that a racist comment?”
So what is different here? Well, on McCain’s ad, he just asks “have you heard about it” and “have you seen it” then asks if you thought it was racist. I have no idea if the 3rd question was asked of the 69% who said yes to the first question, the 63% asked in the second question, or all 1000 people (or even if the 2nd question was just asked to the first 69% or all 1000). Perhaps the answers are in the crosstabs which you have to pay for to see.
However, if the 3rd question is asked of all 1000 people in the survey, then it is quite likely that very few people who haven’t seen the ad would label it as racist since…they haven’t seen it. That would tend to put 26% of respondents into saying “no it wasn’t racist” right off the bat. If he indeed to ask all 1000 people the 3rd question, the 22% who thought it was racist would make up 35% of people who saw it. Not a majority, but not as low as 22%. (I should also note that “Yes” and “Not Sure” put together would make up 59% of those who saw the ad, which is just 2% off of the “Yes” and “Not Sure” totals for the Obama question.)
If he only asked people who heard about the ad, that means at most 690 were asked the question, which would increase the margin of error from 3% to closer to 4%. Not major, but I thought it was ambiguous.
However, the worse part of this poll is that he treated Obama’s comment totally differently. He quoted the one part of his comment, without context, and asked it was racist. He didn’t do like he did with the McCain ad and asked people if they had heard it, and if so, if they thought it was racist, he just gave them a paraphrase of the comment and asked.
And of course, context is important here, since he said it along with a string of other things which are not unimportant. I have a feeling if Rasmussen gave the full quote, not as many peole (which it was 53% in the poll) would have found it racist. Also, Rasmussen explicitly said that “Barack Obama said his opponent will try to scare people” even though the vast majority of the time Obama has used this phrase, he has used the much more vauge “they.”
Also, since Rasmussen didn’t ask how many people had heard the comment before, we have no way to know how many people might have thought it was racist purely from Rasmussen’s reading of what Obama said.
What Rasmussen probably should have done to be consistent is either:
1) Give Obama’s comment and describe the ad (basically how he asked the Obama question, but do it for both)
2) Ask whether they saw or heard the item, and ask if they thought it was racist (how he asked the McCain question, but do it for both), or
3) Simply ask if they think if Obama or McCain have made any racist comments or ads recently and not even mention Obama’s comment or the McCain ad (or by prefacing the questions with something like “there has been an debate about using race between the two campaigns in the past week.”). This is actually probably the best measure of what people think since it takes away the asking of a possibly leading question.
