|
|
May 20, 2013
|
|
|
|
 Going Poll CrazyPosted on Oct 22, 2008
Some polls show Barack Obama with a double-digit lead while others have John McCain even or ahead. Take Pennsylvania, where both Obama and McCain are waging much tougher campaigns than one would expect in light of an 11-point average margin. That’s because their internal polls show a much closer race. So how do you make sense of it all? The short answer is: you can’t. While it’s easy to disparage polls, the science of statistics is not a form of mysticism. Still, the sheer quantity and inconsistency of polls can drive a person mad. Here are a few thoughts on sorting out the mess. Forget the popular vote: National polls are the least important. Presidents are chosen by electors, so if you want to know how Obama and McCain are doing, look at how they’re polling in, say, Colorado and Ohio, not across the board. Not all polls are created equal: Pollsters aren’t just asking a random group of people for whom they’re going to vote, they’re making decisions about whom to talk to and how to balance their responses every step of the way. For example, some pollsters recently started calling cell phone numbers to keep up with changing demographics. These decisions greatly impact the polls. Check out fivethirtyeight.com’s pollster ratings to stay on top of the pollsters’ track records. This election is unique: We’ve never had a nonwhite major-party nominee before. There are record-breaking voter registration drives across the country. YouTube didn’t exist in 2004. There are myriad forces affecting this election that are difficult to account for by looking to either political history or polls. Good news can be bad news: The campaigns depend on a delicate balance of fear and enthusiasm to drive donations and voter turnout. The idea that Obama is somehow going to sweep the election is bad news for his campaign, and the notion that McCain is out of it is bad for his. When people are uncertain or anxious about a decision, they like to see what others are doing. It’s called social proof. It’s human nature to want to know how others are voting, but that shouldn’t stop citizens from analyzing the issues and then following their own consciences. The worst thing about polls is not that they’re often wrong, but that they interfere with a voter’s democratic instincts. Advertisement TAGS:
|
What Will Tighter Restrictions on Trade in Iran Do?
Email 
Print
Share
Email to a friend
Newsletter 
Like Us
Follow Us
Get Our Feed
Tags|Bazaar|Podcast|About Us|Contact Us|User Agreement|Privacy Policy|FAQ: Comments and Moderation | Google+
© 2013 Truthdig, LLC. All rights reserved.
By Gawd, October 23, 2008 at 5:41 pm Link to this comment
(Unregistered commenter)
Well, I don’t think there has been a great deal of change in cellphone-polling issues since the primaries earlier this year. Yet the national polls showed clear evidence of way over-stating support for Obama during the primaries.
Most people forget that Obama regularly led Clinton by 6-10 points or more in the national polls going back to the earliest days of the primaries in February (although she regularly outperformed him vs McCain). However, he never led her in actual votes by more than maybe 0.4%. In fact, the final count of actual votes had her with almost 300,000 more votes than he got while he was still showing a 6-10 point lead over her in the polls.
Only if you gave Obama imaginary or estimated votes from the caucus states that were never cast could he show a lead over her in votes. And then it was only by that extremely narrow 0.4%.
So, if the young voters who have unlisted cellphones weren’t being reached or factored in by pollsters back in February, March, April, May or early June…while the polls were WAY overstating Obama’s support vs Clinton…then Obama is in even bigger trouble than any previous Bradley Effect might suggest.
Report thisBy Mateo, October 23, 2008 at 1:30 pm Link to this comment
(Unregistered commenter)
I still can’t seem to figure out why the media hasn’t really caught on to the fact that the polls are probably going to be more scewed this year than in previoius elections in part because most people under the age of 30 years old don’t own a land-line phone. How are polls conducted? By calling people at their homes, usually at dinner time. I’m here to tell you, the people under the age of 30 are out with their cell phones, the same phones that are not listed in phone books or phone banks. Those who do have land lines screen their calls with caller ID and don’t answer 800 numbers and out of area calls. The people getting polled are older conservatives. Now I admit that older people are more likely to vote than their younger counterparts, but I’ll submit to you that there is an entire demographic being overlooked in the polling. Younger people. So the question is who’s guessing to see what candidate the younger generation will be voting for. I’ve got some guesses of my own.
Report thisBy Gawd, October 23, 2008 at 11:58 am Link to this comment
(Unregistered commenter)
Not true. This is a popular misconception.
In fact, the major national polls have a history of being VERY accurate to the final outcome of the contest while the state-by-state Electoral Map polls have a history of being abysmally inaccurate.
Just a few days before the presidential elecion in 2004, here is what the state-by-state Electoral Map polls were telling us:
Nov. 1, 2004: Kerry_298, Bush_231 (with 270 needed to win)
http://www.electoral-vote.com/evp2004/Pres/Maps/Nov01.html
On the other hand, with extremely rare exception (Fox News Channel’s poll was a notorious exception), none of the major national polls had Kerry ahead of Bush in the weeks before the election and most of them had him behind Bush within a normal Margin of Error to the final voting result.
The most accurate major national poll in 2004 was IBD/TIPP. They got it right to within a fraction of 1 point.
Here is what IBD/TIPP is reporting today. Obama has a 1.1 point lead over McCain:
Report thishttp://www.ibdeditorials.com/Polls.aspx?id=309635713550536
By skulz fontaine, October 23, 2008 at 11:15 am Link to this comment
Recent polling conducted by Okie-Dokie Polling Inc. concluded that fully 88% of all Americans never polled prior by anyone, believe that polling concerns are full of crap. An additional 81% of all Americans never polled prior by anyone, believe that the “big polling concerns are making this shit up as they go along.”
Report thisWell there you go and supported by “staggering polling numbers.”
By GW=MCHammered, October 23, 2008 at 9:20 am Link to this comment
(Unregistered commenter)
Media creates controversy any way it can. Why? Conflict draws viewers and so sells commercials. Trouble is the media believes it can create its own reality. But it’s Bu$hCo Trickle Down Sleaze Reality.
Report thisBy johnny plumber, October 22, 2008 at 8:52 pm Link to this comment
(Unregistered commenter)
i don’t believe in polls; i don’t believe in race; i don’t believe in kennedy; i don’t believe in kings; i don’t believe in barby dolls; i don’t believe in mccains; i don’t believe in democracy; i don’t believe plutocracy; i don’t believe in oligarchy; i don’t believe in moneybags; i don’t believe in absolutes; i don’t believe in karma; i don’t believe in polls (did i say that?); i don’t believe in security; i don’t believe in honesty; and i don’t believe in the ability of the american hedgeocracy to possibly control the most important u.s. election in the past 76 years. i fear for the republic. Just watch your selves and vote on or before Nov.4. May the gods or god guide and protect our last, best hope - Senator Barack Obama.
Report this.....61 year old blue-eyed Nevadan.