On the Inexplicable Oversampling of Democrats
We talked about that Quinnipiac Poll that appeared to seriously oversample Democrats. Hugh Hewitt had Peter Brown, assistant to the director of Quinnipiac Polls and asked him about it:
HH: Do you expect Democrats, this is a different question, do you, Peter Brown, expect Democrats to have a nine point registration advantage when the polls close on November 6th in Florida?
PB: Well, first, you don’t mean registration.
HH: I mean, yeah, turnout.
PB: Do I think…I think it is probably unlikely.
HH: And so what value is this poll if in fact it doesn’t weight for the turnout that’s going to be approximated?
PB: Well, you’ll have to judge that. I mean, you know, our record is very good. You know, we do independent polling. We use random digit dial. We use human beings to make our calls. We call cell phones as well as land lines. We follow the protocol that is the professional standard.
HH: As we say, that might be the case, but I don’t know it’s responsive to my question. My question is, should we trust this as an accurate predictor of what will happen? You’ve already told me there…
PB: It’s an accurate predictor of what would happen is the election were today.
HH: But that’s, again, I don’t believe that, because today, Democrats wouldn’t turn out by a nine point advantage. I don’t think anyone believes today, if you held the election today, do you think Democrats would turn out nine percentage points higher than Republicans?
PB: If the election were today, yeah. What we found is obviously a large Democratic advantage.
HH: I mean, you really think that’s true? I mean, as a professional, you believe that Democrats have a nine point turnout advantage in Florida?
PB: Our record has been very good. You know, Hugh, I…
HH: That’s not responsive. It’s just a question. Do you personally, Peter, believe that Democrats enjoy a nine point turnout advantage right now?
PB: What I believe is what we found.
HH: Geez, I just, and an eight point in Ohio? I’m from Ohio. Democrats haven’t had an eight point advantage in Ohio since before the Civil War. I mean, that just never happens, but Peter, I appreciate your coming on. I’m not persuaded that Quinnipiac Polls haven’t hurt themselves today, but I appreciate your willingness to come on and talk about it.
Isn't that odd? I mean, many observers thought the oversampling -- or should I just say sampling -- was a problem. But if you're the company doing the polling, shouldn't you be able to at least explain it? I'm not saying people would buy your argument, but isn't it weird that there is no defense of it? No attempt to defend it? A friend says that the live radio segment included much more hemming and hawing than is reflected in the transcript.
I think Quinnipiac is a solid pollster, but if they had a reason for this sampling mix, they should be able to provide it. If they didn't, they shouldn't have released a poll with that sample. Right? What is going on?
(h/t Jonah)
- Comment (35)
- · Quote
- · UnfollowFollow (3)
- Pages:
- 1
- 2
Comments:
Jun '12
Re: On the Inexplicable Oversampling of Democrats
It's very simple, Mollie.
The reason is, "we want Obama to win."
But he can't say that to Hewitt, so he has to dance and weave.
May '12
Re: On the Inexplicable Oversampling of Democrats
The latest Pew poll is even worse. It's dem +12. It really is outrageous. I suspect the samples will be adjusted once the election gets closer so that the final polls are more accurate. Right now the pollsters are more worried about getting their poll noticed than having their poll be correct. If they get the final poll in November correct, they know they can still maintain their reputation for accuracy.
Edited on August 3, 2012 at 4:22pmMay '12
Re: On the Inexplicable Oversampling of Democrats
I think the pollsters have an interest in making the race look close or even showing that the weaker candidate is doing well because that keeps hope alive for the weak candidates supporters and keeps them looking at the polls. If the polls consistently show a solid advantage for one candidate, people stop paying attention, and people stop noticing the polls.
Mar '11
Re: On the Inexplicable Oversampling of Democrats
ConservativeWanderer: It's very simple, Mollie.
The reason is, "we want Obama to win."
It is, indeed, pretty simple.
Jul '10
Re: On the Inexplicable Oversampling of Democrats
Sorry, honest polling is a myth. Joe six pack just sees the poll numbers and the oversample is buried or ignored. Mr. Hewitt is clearly fighting the good fight, and I applaud him, but this is exactly the kind of tactic that works well on the uninformed and poorly educated. Or as Obama refers to them, the Base.
May '11
Re: On the Inexplicable Oversampling of Democrats
Handy spreadsheet to reweight any national or state exit poll based on past D/R/I splits. I usually average between 2010 and 2008, which seems like a reasonable projection.
Dec '10
Re: On the Inexplicable Oversampling of Democrats
Pollsters will get work no matter who is in power. There has to be some other explanation for such platant skewing of the numbers. It's a bit conspiratorial, but I wonder how much informal communication is happening between the IRS and polling organizations.
Feb '12
Re: On the Inexplicable Oversampling of Democrats
I think that is probably right, and that motivation will make the data, the methods used to collect them, and the validity of their interpretation a secondary concern (at best). Otherwise, what would be wrong with simply saying, "These are the data we collected. As with any poll, care must be taken in interpreting the results. A voter's enthusiasm or motivation to vote is difficult to assess. Even a person's past track record of participation may not adequately capture who is likely to vote in THIS election. Our poll did not take those factors into account, so the data should be evaluated with that in mind. In the past, our polls have a good track record for predicting electoral outcomes."
Mr. Brown's response does indicate that methodology is of secondary (at best) importance.
May '12
Re: On the Inexplicable Oversampling of Democrats
It's called Self-Affirmation.
You invite to the party/conversation/poll people you already know agree with you, ask questions you already know their answer to, and lo and behold, your beliefs are Reaffirmed.
It's as easy a Sin to fall into as Envy or Covetousness. All you have to be is unsure and incurious enough to not want to risk seeing your beliefs and opinions challenged.
People who repeat, "But I am biased." over and over again are a perfect example of the type who will do this.
Apr '11
Re: On the Inexplicable Oversampling of Democrats
Obama is a wrecking ball and he is taking down everyone one within reach. These pollsters have to fudge or obfuscate the numbers to make people believe Obama is still loved, that Hope and Change is still alive and well, that he is still The One. Not that these guys weren't teetering already but no one is going to believe them after the bald-faced fraud that they are perpetrating now.
Jun '12
Re: On the Inexplicable Oversampling of Democrats
Edward Smith: It's called Self-Affirmation.
You invite to the party/conversation/poll people you already know agree with you, ask questions you already know their answer to, and lo and behold, your beliefs are Reaffirmed.
It's as easy a Sin to fall into as Envy or Covetousness. All you have to be is unsure and incurious enough to not want to risk seeing your beliefs and opinions challenged.
People who repeat, "But I am biased." over and over again are a perfect example of the type who will do this. · 9 minutes ago
Nail. Head.
Jul '10
Re: On the Inexplicable Oversampling of Democrats
Pollsters roll right over to voicemail in the Sisyphus household, and we don't qualify to be a Nielsen family, either. We don't answer questions about race or religion or income level, for starters, so there's not much point. It is just as well. Pollsters being the enemy of democracy, when we do talk to them we lie. This drives our Lefty acquaintances insane. Apparently it is a public duty to give honest answers to dishonest questions in their view.
And, of course, biased by definition means "I dismiss your views because they are not mine."
Oct '10
Re: On the Inexplicable Oversampling of Democrats
I think this really is the answer. Can he really be so ideologically committed that he couldn't say this, or is he worried that admitting this too clearly would undermine the sales of his product, or is he just unable to think on his feet and be as articulate as radbiochem?
Jun '11
Re: On the Inexplicable Oversampling of Democrats
He's channeling Nigel Tufnel: "The numbers all go to eleven."
Aug '10
Re: On the Inexplicable Oversampling of Democrats
Let's see what they do over at RCP ?
Has anyone at Rico ever been polled ? Who is home in the middle of the afternoon when they make these calls ? Employed people ?
It's like speaking to a child. "But dad everybody has one !" Not wanting to miss out being in the mainstream, Obama's pollsters are setting up a facade of popularity .
I am going to pull out my Canetti and reread the crowd psychology stuff. Careful with that link , Clive James can keep your attention a very long time.
Mar '12
Re: On the Inexplicable Oversampling of Democrats
The sample that counts is in November.
I have never been polled and I have voted in every presidential election since I was of age, that being the Carter election, so I am a voter who takes this seriously. I took it seriously when I did not think (Carter) and took it seriously when I was transformed and voted for Reagan.
I've bee serious ever since. Did not always win, but serious none the less.
Oct '10
Re: On the Inexplicable Oversampling of Democrats
"Our record has been very good, our record has been very good, our record has been very good."
Yeah, and the Yankees' record is very good. But sometimes they lay an egg. So, Peter, did you lay an egg with this particular poll? Not with your entire reputation for polling, but with this one particular poll?
What a weasel. What an abject weasel. When a man cannot admit to a mistake or being wrong, he's an untrustworthy weasel.
Dec '11
Re: On the Inexplicable Oversampling of Democrats
ConservativeWanderer: It's very simple, Mollie.
The reason is, "we want Obama to win."
But he can't say that to Hewitt, so he has to dance and weave. · 1 hour ago
Why couldn’t he say that?
In all honesty here, I’m not trying to be provocative, but in today’s climate, on a conservative talk radio program, what is the down side to saying: “Hugh, we designed this poll with the intent of it serving as an endorsement of the President”?
Who would be influenced by that away His Majesty?
I can almost guarantee the Media Matters types who listen in would pump their fist in exaltation, it would certainly motivate the base, which all that the King is pandering to now.
Sep '10
Re: On the Inexplicable Oversampling of Democrats
radicalbiochemist
"These are the data we collected. As with any poll, care must be taken in interpreting the results. A voter's enthusiasm or motivation to vote is difficult to assess. Even a person's past track record of participation may not adequately capture who is likely to vote in THIS election. Our poll did not take those factors into account, so the data should be evaluated with that in mind. In the past, our polls have a good track record for predicting electoral outcomes."
I think the above goes without saying, which is why Mr. Brown didn't say it. They followed their standard methodology of random phone sampling, including mobile numbers. This does not assess which of these voters is likely. That's a difficult task that they're not taking on. In many elections, that omission will result in errors.
I believe that Rasmussen tries to poll likely voters and have some other innovative features. They use machines instead of human operators. How this affects results is controversial; there's a case to be made that it is advantageous.
No need for conspiracy or other exotic theories. Take their results at face value, with the usual caveats.
Jun '12
Re: On the Inexplicable Oversampling of Democrats
Bluenoser
In all honesty here, I’m not trying to be provocative, but in today’s climate, on a conservative talk radio program, what is the down side to saying: “Hugh, we designed this poll with the intent of it serving as an endorsement of the President”?
Who would be influenced by that away His Majesty? · 2 minutes ago
What website or media outlet, wanting to appear unbiased -- and there are a lot of them -- would commission a poll from a polling firm that admitted to skewing the polls?
It's a matter of business. If he admits skewing, he loses business.
Also, on the ideological side, if he admitted skewing the poll, every single poll out of Quinnipiac would be greeted by conservatives -- including me -- reminding readers that they publicly admit to skewing polls.
In other words, admitting it eviscerates his effectiveness.