Was Iowa Poll Finding Hillary Ahead Skewed By The Holidays?
That new ARG survey that Drudge has been pushing all day that finds Hillary with a double digit lead over Obama in Iowa -- was it skewed by the fact that it was taken from Thursday through Sunday, which is to say, on the weekend before Christmas?
Mark Blumenthal over at Pollster.com says that we should approach the poll with extreme caution -- and gives us lots more on the historical difficulties of polling during the holidays as an added bonus.
Advertisement















I think we all understood that this double digit lead is nearly impossible for any candidate on the Democratic side.
December 26, 2007 5:32 PM | Reply | Permalink
yeah, that seems hard to argue with.
December 26, 2007 5:35 PM | Reply | Permalink
You must be naive to believe in the last ARG poll, politico.com reports that reporters in IA are suspicious of the last ARG poll, they believed that it was skewed due to numerous reasons. Check out at drudge web, he called it "shock poll." Sorry for using drudge, I don't like him either but it is worthy noticing that many people, besides Hillary supporters, had doubt about this poll.
December 26, 2007 5:36 PM | Reply | Permalink
I work for a research firm (consumer goods research, not voter polls) -- we stopped running research on Thurs, Dec 19 and will not start again until Mon, Jan 7. Perhaps polling voters is different or they have a better way to account for missing people, but we would never rely on results conducted over the holidays.
December 26, 2007 5:40 PM | Reply | Permalink
I kind of thought the Obama number looked a little low and Clinton was probably a tad high in this one. ARG had done another poll just a few days previously that had Obama in second place and Clinton in the lead by a much smaller margin.
However this is when you would expect to start seeing the effect of personal contacts from field operations getting out there turning trying to harden their "twos" into "ones." So we could be picking up some of that. Edwards and Obama have also started kind of going after each other in the last week or so, and the colleges were probably on holiday break when the second one was done.
Anyway, I'll be very surprised if we don't see at least a couple more polls in the next week. And of course in 8 more days we can stop guessing.
December 26, 2007 5:40 PM | Reply | Permalink
John Edwards must be leading in Iowa since the media (Hardball) has resorted to showing "RUNNING AVERAGES" of Iowa poll numbers, as if "AVERAGES" show current movement in the polls.
An "AVERAGE" by definition, takes in the positive and negative movement, so it's pretty much telling you NOTHING about now, as if Hardball doesn't know that.
But, those numbers make it look like Edwards is doing worse right now in Iowa, so of course those are the numbers the media is going to show.
Earlier today I sent a scathing email to my local Newscorp/Viacom affiliate here in North Carolina (WRAL 5) for showing those FLUKE ARG Iowa poll numbers that you are talking about, not because they showed them, but because they STILL haven't shown, ON THE AIR, those Public Policy Polling numbers that show Edwards would beat any Republican in North Carolina, nor have they even mentioned the CNN Gallup Poll from a few weeks ago that showed Edwards beating all Republicans. This from WRAL, which has also simulcasted news coverage with CNN, which means they are affiliated with it as well.
Even the LOCAL CORPORATE MEDIA IN NC, is doint all that it can to marginalize Edwards.
They show his bad numbers (this fluke ARG poll), but they ignore his good numbers.
December 26, 2007 5:59 PM | Reply | Permalink
One thing for sure that HRC has managed to stop the bleeding in last couple of weeks. It looked really bad for HRC early Dec. I thought after Oprah week and many snafus that HRC campaign made, Obama might sweep Iowa. Somehow may be Bill Clinton effect, Michael Wholey's startegies or softening of her image the race is still very tight. You cannot rule out anybody. It will be really bad for Obama if Obama comes in 3rd as the main stream media has raised expectation for him to win Iowa. Key for him is turnout of voters between 18-34. Obama has potential to win big or lose big.
December 26, 2007 6:02 PM | Reply | Permalink
Yeah, you did, but not the people who THE TELEVISION MEDIA showed it to.
December 26, 2007 6:10 PM | Reply | Permalink
Pandora - I agree with you. Clinton has stopped the bleeding (I think the DMR endorsement helped quite a bit - psychologically for her team, if nothing else), and Obama could come in a high first or a distant third. It's impossible to tell which at this point. Perhaps it's because I'm a supporter, but my nerves tell me either scenario is "highly likely."
December 26, 2007 6:37 PM | Reply | Permalink
We will see, maybe its just a sign of things to come. The last ARG and Rasmussen were almost identical, showing HRC pulling ahead of BO by 4 point or so.Its funny how the republican polling on the same poll was not'skewed' or whatever, isn't it.
December 26, 2007 7:04 PM | Reply | Permalink
Yes it was skewed.
It included more 1st-time caucus goers, which was of course designed to hurt Edwards who is strong among former caucus goers.
Anyone who tells you Edwards won't win doesn't know what they're talking about. All the experts like Yepsen and Halperin know Edwards will win.
December 26, 2007 7:50 PM | Reply | Permalink
As best I can figure -- I finally broke down and started charting the race for myself -- Clinton actually flattened out the first week in December and has been trending ever so gradually upward since then. I have her back at her early November high of 30% right now, including these two latest ARG polls or 28-29% without them.
Obama started flattening off at about 30% around the 10th of the month in my analysis. Right now he's a tick under 30% without the two ARG polls or three points lower with them.
Edwards is at 22-23% for the last month, with or without the two ARG polls.
December 26, 2007 7:58 PM | Reply | Permalink
Oh, BTW Pandora, Michael Whouley apparently isn't in Iowa as it turns out. But Teresa Vilmain is, and she's supposed to be among the people who are in a class as Whouley. I still have my suspicions that the helicopter may still have been Whouley's idea though. That seems to be one of his favorite stunts.
December 26, 2007 8:25 PM | Reply | Permalink
The change in the "big mo" in favor of Hillary was palpable about 2 weeks ago, and I had pointed it out. She is now positioned to do very well in IA, because of the overall positive turn of events for her and not because of this poll ARG poll. This poll is as reliable as every other poll that's been taken in IA: Regardless of the numbers and screaming headlines, no one has ever had clue about who was ahead in IA... That remains true right now, even after the ARG poll.
The most sensitive index of overall change in "big mo" is the INTRADE Prediction Markets (see on the right). It nicely tracked the Obama surgelet and has now already picked up tge latest dynamics: Obama has stalled and Hillary is regaining some of her previous edge...
December 26, 2007 8:40 PM | Reply | Permalink
December 26, 2007 9:18 PM | Reply | Permalink
Who is kidding who ? Every time the polls show Hilliary ahead, some sceptic claims there is some abnomily at work. Maybe she is surging !
December 26, 2007 9:25 PM | Reply | Permalink
I already knew holiday polling was considered unreliable by the professionals, but it was indeed quite useful to get the details from Blumenthal insofar as they helped explain why Clinton's numbers in particular would benefit and Obama's numbers in particular would be harmed.
By the way, pollster also has a great analysis of the trends in Iowa based on varying the sensitivity in their regressions and noting the general problem with different baskets of polls coming in at different times. The bottomline is that the Clinton supporters above may be right insofar as Clinton may have reversed her downward trend, but it could also be the case that what they are seeing is an artifact of random noise and methodological variance.
And the bottomline is we probably won't know which theory is right until after the caucuses, thanks in part to the holiday effect, and in part to the general closeness of the race.
December 26, 2007 10:06 PM | Reply | Permalink
@dschungu and DemAC:
Thank you for your informed and moderate input. I sometimes post just to pour gasoline on the crazies, but always appreciate your soothing presence(s).
December 26, 2007 10:37 PM | Reply | Permalink
Mark Penn, Hillary strategist, union buster and Blackwater consultant when had his friends at ARG releasing their outlier poll on Christmas eve reminds me of this guy in GA who went to deposit a 1million dollar bill at his local bank thinking that people were stupid enough to realize that it was fake. Spin all you can about polls, outcome in IA will depend on turnout, organization and numerous other variables. As a pragmatist, I will say let's just wait and see. Anyone remember where Kerry was at this day in the last election?
December 27, 2007 1:01 AM | Reply | Permalink
Now it comes down to who controls the offensive and defensive lines.
December 27, 2007 6:03 AM | Reply | Permalink
DemAC and dcshungu are Clinton attack dogs.
NOTHING they say has an ounce of credibility with anyone except Hillary's spin machine.
They could save much time and space by simply saying "See our previous posts" and "ditto".
December 27, 2007 7:05 AM | Reply | Permalink
MonaL spins: "@dschungu and DemAC: Thank you for your informed and moderate input. I sometimes post just to pour gasoline on the crazies, but always appreciate your soothing presence(s)."
This is too funny. Did someone else say the term "dishonest pittbulls" comes to mind.
More of Hillary's corporate paid operatives give each other cover. If you have not substance, the only alternative is to lie and spin. Something like: "vote for Hillary, the most experienced candiate." We'll soon see if Iowa voters fall for that bull.
December 27, 2007 7:13 AM | Reply | Permalink
Thanks to TPM for posting the link to Charles Franklin's analysis of the ARG polls. http://www.pollster.com/blogs/pollster_variation_in_iowa.php
Bottom line, ARG polling overestimates Clinton's support and underestimates Obama's.
Election Central. Some posters claim an ARG connection to Marc Penn? Is this true? Please post a fair response.
If there is question about the crediblity of ARG, why does EC present it without qualification? Evidently, because it's purpose is to give Hillary a boost.
December 27, 2007 7:46 AM | Reply | Permalink
Obama's numbers may have been low because of college students being gone for Christmas.
December 27, 2007 8:54 AM | Reply | Permalink
Mark Blumenthal over at Pollster.com says that we should approach the poll with extreme caution --....why might it blow up..lololololololol what a joke. Might it tear a hole in the fabric of the universe.
December 27, 2007 10:29 AM | Reply | Permalink