Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

kentuck

(111,107 posts)
Fri Oct 26, 2012, 07:27 PM Oct 2012

Why is the Gallup Poll so at odds with other polls??

The latest Gallup poll says Romney is ahead by 5 points in a nationwide poll of "Likely Voters" - something like 50% to 45%. How do they get those numbers?

Do they poll cell phones? Do they poll door to door? Do they call people that haven't yet voted? Or is it at random?

Either way, I have my doubts about their process. At one time, they were a well-respected polling organization. If they are off this much from everyone else, they could lose all credibility, if they are wrong?

From what I have seen and read, President Obama has been leading in almost all the early voting. In some cases, by large margins. This is a great start if we can get people to the polls a week from this Tuesday?

My suggestion is that we ignore these polls. All of them. We should simply focus on getting out the vote, right up until and including election day. We should stay positive and smile every time you pass a Republican. It will drive them crazy.

18 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies
Why is the Gallup Poll so at odds with other polls?? (Original Post) kentuck Oct 2012 OP
October 26, 2000. Bush 52%, Gore 39% scheming daemons Oct 2012 #1
They're the only ones being sued by the Obama Admin MjolnirTime Oct 2012 #2
Nate Silver wrote a great article on just this topic BlueState Oct 2012 #3
but nycbiscuit Oct 2012 #16
Arrogrance. They continue to way oversample the soft Thrill Oct 2012 #4
I cannot believe you have a record of over 63,000 posts on DU... DonViejo Oct 2012 #5
I do apologize. kentuck Oct 2012 #10
DOJ lawsuit timlot Oct 2012 #6
Gallop finished 17th of 20 polls for accuracy in 2008. bluestate10 Oct 2012 #7
Not correct. Gallup includes cell phones in their sample. tritsofme Oct 2012 #9
They will make a major correction nevergiveup Oct 2012 #11
In my opinion, the owners of Gallup have a vested interest Live and Learn Oct 2012 #8
It's designed specifically to drive DU poll-hounds crazy Bucky Oct 2012 #12
You are, by far, the best at what you do alcibiades_mystery Oct 2012 #13
according to Simon Jackman (Huffpo Pollster) InsultComicDog Oct 2012 #14
That's an interesting graph! kentuck Oct 2012 #15
"House effect" InsultComicDog Oct 2012 #17
the "likely voter" concept is a model, not an absolute 0rganism Oct 2012 #18

BlueState

(642 posts)
3. Nate Silver wrote a great article on just this topic
Fri Oct 26, 2012, 07:34 PM
Oct 2012
http://www.nytimes.com/2012/10/20/us/politics/in-national-polling-its-gallup-vs-the-rest.html?_r=0

He demonstrates that in the past they have been inaccurate.

His conclusion is that it is their methodology. Specifically the likely cause is their methodology in determine who is likely to vote.

It's a good read for anyone (probably alot of us here) who have become addicted to polls. It might be wise to
avoid them until the election as you recommend but I doubt I will be able to.

nycbiscuit

(46 posts)
16. but
Fri Oct 26, 2012, 08:24 PM
Oct 2012
http://www.mediaite.com/online/gallup-rasmussen-forecast-more-republican-voters-in-2012-than-2004/

It would seem Gallup and Rasmussen were close on forecasting turnout in previous elections, and for this election the numbers don't look good for Obama.

I do take this with a grain of salt, considering they have both been leading shamelessly to the right in their polling.

Thrill

(19,178 posts)
4. Arrogrance. They continue to way oversample the soft
Fri Oct 26, 2012, 07:35 PM
Oct 2012

Even when they've been called on it. They refuse to change it

DonViejo

(60,536 posts)
5. I cannot believe you have a record of over 63,000 posts on DU...
Fri Oct 26, 2012, 07:36 PM
Oct 2012

and have, apparently, never seen this topic discussed here. I've seen at least ten posts like this in the last week.

bluestate10

(10,942 posts)
7. Gallop finished 17th of 20 polls for accuracy in 2008.
Fri Oct 26, 2012, 07:49 PM
Oct 2012

2008 marked a watershed because during that election cellphones were becoming popular. Cellphones are even more prevalent today morphed into smartphones that email and surf the web. The groups supporting Obama do most of their computing and communication over smartphones, Gallop is not reaching that demographic.

nevergiveup

(4,763 posts)
11. They will make a major correction
Fri Oct 26, 2012, 07:57 PM
Oct 2012

24 hours before the polls open. They are a major player in the "horse-race" hype that is permeating the press.

Live and Learn

(12,769 posts)
8. In my opinion, the owners of Gallup have a vested interest
Fri Oct 26, 2012, 07:51 PM
Oct 2012

in seeing Romney elected and are doing their part to help him steal the election. Just look who owns Gallup and what they do.

http://sri-2000.com/company.html

Selection Research International, Inc. (SRI)

Selection Research International, Inc. (SRI) specializes in evaluating employees and their families for international assignments and building global talent development systems.

<snip>

For maximum effectiveness, assessing employees and their families for international assignment should be integrated into a larger global talent development system. SRI begins all new engagements with a one-day working session, structured to integrate SRI tools and processes into existing and planned talent management programs of the corporation.


Bet Romney already supplies them with a ton of business.

Bucky

(54,041 posts)
12. It's designed specifically to drive DU poll-hounds crazy
Fri Oct 26, 2012, 07:58 PM
Oct 2012

And give the Anti-Concern Troll Counterterrorism squad something to kvetch about.

 

alcibiades_mystery

(36,437 posts)
13. You are, by far, the best at what you do
Fri Oct 26, 2012, 07:59 PM
Oct 2012

You have my unflagging admiration.

Seriously, and you know what I'm talking about. You should really receive an award. What you've managed over ten plus years is an astounding achievement.

InsultComicDog

(1,209 posts)
17. "House effect"
Fri Oct 26, 2012, 11:38 PM
Oct 2012

the amount of points off of the average, on average, that each pollster is, ordered from most pro-Obama to most pro-Romney.

0rganism

(23,962 posts)
18. the "likely voter" concept is a model, not an absolute
Fri Oct 26, 2012, 11:50 PM
Oct 2012

it's not like there's a pure reservoir of "likely voters" for Gallup to poll. rather, they're mixed in with the registered voters, and the pollsters have to figure out "likely" from self-reported demographic and historical information. If a certain set of criteria are satisfied, the respondent is ranked as a "likely voter". Frequent deviations from the rest of the polls probably stem from differences in this set of criteria, plus whatever other weighting factors the nice folks at Gallup saw fit to apply to the responses.

Latest Discussions»General Discussion»Why is the Gallup Poll so...