Polling 2012

[Stoked]

https://realclearpolitics.com/epolls/latest_polls/

Look at the two polls ffrom today on Michigan. One has Romney +1 and the other Obama +14. Which one would you be inclined to belive and why?

Also can someone adequately explain why there is such a wide discrepancy? 16 point swings on the same day would indicate more than biased polling right?

 

[PearlWatson]

https://realclearpolitics.com/epolls/latest_polls/

Look at the two polls ffrom today on Michigan. One has Romney +1 and the other Obama +14. Which one would you be inclined to belive and why?

Also can someone adequately explain why there is such a wide discrepancy? 16 point swings on the same day would indicate more than biased polling right?

I looked into PPP...the one who has +14 for Obama in Michigan. The discrepancy could lie in who happened to be willing to take the poll, and sample size.

This is what they have to say about the numbers. The Condi Rice numbers are surprising to me.

There are a couple common themes driving Obama's success in these states. The biggest
is his competitiveness with white voters. He actually leads 48-44 with them in Michigan,
and he just narrowly trails 46-45 with them in Pennsylvania. Given the healthy sized
African American populations in both those states anything close to a split with white
voters will give Obama an easy victory. Obama is also benefiting from a huge lead with
women in both states, 59-33 in Michigan and 54-37 in Pennsylvania, which helps him
offset a closer race with men.
One thing that could be a huge game changer in both of these states? Mitt Romney
selecting Condoleezza Rice as his running mate. She would give him a 6 point advantage
both places, pulling him into a tie in Pennsylvania at 45-45 and narrowing the gap in
Michigan to 50/42. Rice is a very unusually popular political figure. In Pennsylvania her
favorability rating is 60/27 and in Michigan it's 56/28. She's even seen favorably by
Democrats- 47/38 in Pennsylvania and 41/40 in Michigan- in both states. Her selection
has the potential to be a game changer.

https://www.publicpolicypolling.com/pdf/2011/PPP_Release_MIPA_72512.pdf

 

[One Brow]

https://realclearpolitics.com/epolls/latest_polls/

Look at the two polls ffrom today on Michigan. One has Romney +1 and the other Obama +14. Which one would you be inclined to belive and why?

Also can someone adequately explain why there is such a wide discrepancy? 16 point swings on the same day would indicate more than biased polling right?

Polls can be affected by the order of the questions, the nature of the questions, the wording of the quesitons, and even the tone of voice of the person asking the question, plus many other things, without any deliberate attempt at bias.

 

[nightmare3983]

https://realclearpolitics.com/epolls/latest_polls/

Look at the two polls ffrom today on Michigan. One has Romney +1 and the other Obama +14. Which one would you be inclined to belive and why?

Also can someone adequately explain why there is such a wide discrepancy? 16 point swings on the same day would indicate more than biased polling right?

I'm not sure why there is the discrepancy, but I would give more weight to the poll with Obama winning because they showed the exact question they used and provided information about how many people were polled. The other poll didn't have any of that.

 

[One Brow]

Looking at the questions, the difference between "approval" in Q1 and "favorable" in Q2 could easily introduce bias for Obama.

 

[Gyp Rosetti]

I'm not sure why there is the discrepancy, but I would give more weight to the poll with Obama winning because they showed the exact question they used and provided information about how many people were polled. The other poll didn't have any of that.

This. I'd also add that even if the gap isn't +14, I'd guess he still leads.

 

[Scat]

This is interesting as I was reading an article about polling just earlier today. Essentially, when a polling place calls you they typically already know your political affiliation. They can sway a poll simply by who they call, and yes, they know who they're calling.

Also, the way they word questions is very important. The article said that in one poll, when they asked people if they were satisfied or unsatisfied with Obama's performance thus far, 95% of repubs said dissatisfied and 65% of Dems said satisfied. When they added a 3rd option, unsure, repubs stayed a 95% dissatisfied but dems dropped to 35% satisfied with the other 30% choosing unsure. To me that is a HUGE difference and completely changes the forecast of the upcoming election.