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

dsc

(52,162 posts)
Mon Nov 5, 2012, 10:12 PM Nov 2012

It has been repeatedly pointed out that the only way Obama will lose is if the polls are biased

but there is only one reasonable historical way polls have been biased. It has been when you have had a Bradley affect. Well there is only one historical candidate this time, and he is Romney. Romney would be the first Mormon President and many of the GOP's evangelical base consider Mormons to be cultists on par with Hare Chrishnas or Jehovahs Witnesses. If the polls are biased, I think a case can be made that there is a much greater chance that they are biased in favor of Romney than biased in favor of Obama. Yet another reason to be confident in Obama's victory.

9 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies
 

Coyotl

(15,262 posts)
1. The other bias is likely under-representation of Obama support, not Romney = Latinos, youths, poor
Mon Nov 5, 2012, 10:15 PM
Nov 2012

The people likeliest to be left out of polling may well break heavily Obama.

 

Democratopia

(552 posts)
4. This looks likely, especially with Latinos
Mon Nov 5, 2012, 10:23 PM
Nov 2012

Plus over-representation - the Republicans are increasingly dependent on older whites, especially males - how many have passed on in the four years since the last election.

ItsTheMediaStupid

(2,800 posts)
2. Good point - there is that chance that people would say they're voting for him and stay home
Mon Nov 5, 2012, 10:16 PM
Nov 2012

Or write in Jesus, or vote libertarian, but they don't want their neighbors, wife, kids, rednecks down the block to know they didn't vote R in 2012.

 

Democratopia

(552 posts)
3. Very good point OP. And another bias would be
Mon Nov 5, 2012, 10:20 PM
Nov 2012

if suddenly Micky Mouse polling organizations appeared with the intention of making one side look better than it really is.

Thrill

(19,178 posts)
5. I don't get the biased meme. Keep in mind the Polls already are biased against him
Mon Nov 5, 2012, 10:23 PM
Nov 2012

because the majority of the polls are LV models. Where they basically cut half his leads from the RV poll numbers. Because they always assume Black, Hispanic, and young voters won't actually vote

 

Democratopia

(552 posts)
6. And I think that even some of those early voters wouldn't say they are "likely" to vote.
Mon Nov 5, 2012, 10:33 PM
Nov 2012

They go because others are there - friends, family, fellow church members etc.


 

Coyotl

(15,262 posts)
7. And what is the probability so many polls share the same biases? P = Zip
Mon Nov 5, 2012, 10:35 PM
Nov 2012

If they are inaccurate, we should expect a bell curve with equal error in both directions, not a systematic problem with statistics. One problem all pollsters might encounter is inability to phone all segments of the populace equally.

dsc

(52,162 posts)
9. It is possible the model is systematically wrong across polls
Mon Nov 5, 2012, 10:39 PM
Nov 2012

but the likelihood of it is frankly very, very small. In essence Silver says it is about 8.5%, I think it is likely even lower than that.

Latest Discussions»Retired Forums»2016 Postmortem»It has been repeatedly po...