Sunday, October 14, 2012

Why I'll never trust PPP's polls

PPP has a habit of slipping strange push-poll-like questions into their supposedly neutral polls about politics. I was polled last month and asked who deserved more credit for killing OBL: Romney or Obama. When I asked a former student of mine who now works for PPP what that question was all about, he just said that they sometimes ask quirky questions to get at how people are thinking. When I asked why there wasn't a parallel quirky question to get at strange things Democrats might believe, he simply shrugged and said that PPP never claimed not to be in favor of the Democrats. I argued that they certainly did try to claim that when they presented their polls as neutral, unbiased measurements of opinion. He had no answer for that and since he's a nice guy and I have extremely fond memories of him as an 8th grade student and we were talking at a panel in front of present-day students, I let it go. However, his answer has colored any tendency I might have to give credence to a PPP poll. And now here is more such evidence of how PPP slips in their "quirky" question into supposed straight polls from a poll they did in Wisconsin.
I spoke by phone today with Wisconsin voter Dave Summers, who lives in the Madison area. He told me, “I got a survey. I don’t normally answer these calls, but I did [this time]. I started out pretty normal — President, Senate.” However, he said it then got weirder. The automatic survey asked if he had a favorable or unfavorable impression of Sean Hannity and Tucker Carlson. He said he was greatly disturbed when the automated call then asked, “Do you believe conservative media want white people to think Barack Obama hates them?”
That was just like the poll with the OBL question - it started out with straight questions about whom I was voting for and what my opinion was of various figures. And then it slipped in that doozy of a question that seemed to come from nowhere.

Rubin followed up with PPP to ask them about their question.
I called Tom Jensen of PPP. He said that it was his poll. I asked whether this wasn’t a classic advocacy poll designed to get a specific answer. He demurred, “Well, we were asking a series of questions about conservative media.” He said that this call followed the posting on Drudge of the 2007 video in which then-Sen. Barack Obama (D- Ill.) talks about Hurricane Katrina and denying aid to residents. He claimed that since conservative media were trying to make an issue of this (in fact most conservative outlets downplayed or ignored the issue), it was important to see whether that effort (to poison the thinking of white voters, I suppose) was “successful.”

The questions on conservative media and on white people were asked at the very end of the poll.

Wasn’t a question that put the conclusion in the question a leading question? “No, it wasn’t a leading question,” Jensen insisted. He then sent over the questions, which for reasons he did not explain were not released publicly.

But of course the question is one that imparts the information to the voter in the most slanted way possible. There are no questions about liberal media or bias. To ask the question is to assume that conservative media are up to make white people fear Obama

Every independent or Republican pollster I asked was shocked by the poll question, though many declined to talk on the record about the work of a competitor.
Rubin then posts the script of the PPP questions in their Wisconsin poll.

I think that their whole purpose was to get a result and then publish it to say that a certain percent of Republicans think that Obama hates white people and then to follow up with a statistic about how many blame conservative media for pushing the story. They did the same thing with the poll question I was asked - they put out their results and let it get picked up by liberals who could then use it as a talking point against Republicans.

Remember these two examples of how PPP is using their polls to manufacture a storyline that can then be turned back on Republicans. If that is the type of goal that they have in their polling, why shouldn't we be skeptical of the results that they get in their supposedly neutral polling?