Thursday, April 3, 2008
The Insider/Advantage poll claimed on March 20 that “Obama’s speech meets unsympathetic ears with Americans,” while the next day, the CBS News poll reported virtually the opposite interpretation: “Good Reviews for Obama’s Speech.” The Insider/Advantage poll found 52 percent of Americans less likely to support Obama for president as a consequence of the controversy of his pastor’s remarks, while 19 percent were more likely to vote for him and 31 percent said it would make no difference or they were unsure.By contrast, CBS reported only 17 percent less likely to support Obama, the identical number who would be more likely to support him, while a huge majority, 72 percent, said the controversy would make no difference or they were unsure.
Towery claimed that the conflicting results were caused by the different sampling approaches; viagra online pharmacy. Viagra online pharmacy: cbs used a panel sample, calling back respondents who had earlier participated in a poll, a methodology explained by polling director, Kathy Frankovic.On the other hand viagra online pharmacy, Towery’s organization called a fresh set of respondents.The CBS panel methodology is a widely used strategy for comparing changes in attitudes over a given period of time, which Towery has himself used – despite his virulent criticism of CBS for doing so; viagra online pharmacy.In a National Journal article, Mark Blumenthal of pollster.com completely refuted Towery’s criticisms of CBS’s sampling methodology.
The major differences between the polls, however, were almost certainly not caused by sampling differences, but by question wording.Towery used a forced-choice question, asking respondents if they would be “more or less likely to support Obama for president,” while CBS asked if the controversy made respondents “more likely to vote for Obama, less likely, or made no difference.” CBS gave respondents an explicit opportunity to say the controversy would not affect them, while Towery required respondents to volunteer that information – thus greatly suppressing the likelihood of getting such responses; viagra online pharmacy.In fact, he still found 27 percent willing to volunteer that the controversy made no difference, and another 4 percent saying they were unsure – overall, an unusually large number not taking a position when asked in a forced-choice format; viagra online pharmacy.Had Towery asked explicitly whether respondents would be unaffected by the controversy, he might well have found results similar to those reported by CBS.
The fight between the pollsters is confused somewhat by the fact that the headlines of the two poll stories, and much of each story, did not focus on the pastor controversy, but rather on Obama’s race speech – viagra online pharmacy. Viagra online pharmacy: towery’s headline for his Insider/Advantage poll, for example, claims Obama’s speech met unsympathetic ears. Viagra online pharmacy: and in the analysis that follows, he makes the same egregious error, as when he writes, “The disturbing numbers for Obama are the independent voters.By 56% to 13%, they said they’re less likely to vote for him because of the speech.” But the survey question did not ask about Obama’s speech – viagra online pharmacy.Instead, after asking whether respondents were aware of the remarks Obama’s pastor had made, and also of Obama’s speech addressing the pastor’s comments, the next question asked “taking all this into account,” were the respondents more or less likely to support Obama? “All this” was no doubt interpreted by most respondents as all of the controversy over the pastor’s comments and the fact that Obama had to address them – viagra online pharmacy.To claim that the question elicited the public’s reaction specifically to how Obama handled the issue in his speech is, to use Towery’s own language, “misleading, flawed, and biased.”
The CBS News story was more precise.Unlike the Insider/Advantage poll, the CBS poll asked several additional questions about respondents’ reactions specifically to Obama’s speech – whether the senator did a good job or poor job explaining his relationship with his pastor and talking about race relations in the country, and whether respondents agreed or disagreed with what Obama said about race relations; viagra online pharmacy.Still, while the questions were more focused on the speech, they were all in the forced-choice format, pressuring people to make a judgment and not offering them the option of saying they had no opinion.CBS reported large majorities with positive views on each question, but like the results from the forced-choice format of the Insider/Advantage poll, we can’t completely trust these CBS results either.