HomeLatest ThreadsGreatest ThreadsForums & GroupsMy SubscriptionsMy Posts
DU Home » Latest Threads » Forums & Groups » Main » Politics 2014 (Forum) » The time has come to take...

Fri Nov 9, 2012, 01:54 AM

The time has come to take the (D) off of PPP Polling.

One of the frustrations of the last campaign season was the designation of various polling. Gravis was never identified as R even when it issued polls which gave Romney huge leads in the swing states of CO, VA and NC.

PPP on the other hand seemed to have the most accurate polling. It labeled NC as a tie when Gravis was giving Romney an 8+ advantage. PPP is always identified as a Democratic polling firm because they are frequently hired by Democrats. And yet in 2010 they went against the pack in predicting first that Brown would take Kennedy's seat.

Fordham University examined and ranked the various polling. The most accurate pollster was PPP. PPP therefore has to be considered the baseline and from that point all other pollsters you would either fall in the range of error or show a consistent R or D bias. The pollster who is identified as the most accurate cannot be categorized as having a bias, by definition. PPP is the only pollster that cannot be given either a D or R.

The fact that the most accurate pollster is usually hired by Democrats, or that they are run by Democrats, or that they make the most sense to Democrats doesn't mean that PPP leans Democratic it means that Democrats lean to objectivity.

The media should stop attaching 'D' to PPP, they are only embarrassing themselves.





http://www.fordham.edu/images/academics/graduate_schools/gsas/elections_and_campaign_/poll_accuracy_2012_presidential_election_110712.pdf


Fordham Ranks the accuracy of various pollsters.


1. PPP (D)*
1. Daily Kos/SEIU/PPP*
3. YouGov*
4. Ipsos/Reuters*
5. Purple Strategies
6. NBC/WSJ
6. CBS/NYT
6. YouGov/Economist
9. UPI/CVOTER
10. IBD/TIPP
11. Angus-Reid*
12. ABC/WP*
13. Pew Research*
13. Hartford Courant/UConn*
15. CNN/ORC
15. Monmouth/SurveyUSA
15. Politico/GWU/Battleground
15. FOX News
15. Washington Times/JZ Analytics
15. Newsmax/JZ Analytics
15. American Research Group
15. Gravis Marketing
23. Democracy Corps (D)*
24. Rasmussen
24. Gallup
26. NPR
27. National Journal*

30 replies, 2244 views

Reply to this thread

Back to top Alert abuse

Always highlight: 10 newest replies | Replies posted after I mark a forum
Replies to this discussion thread
Arrow 30 replies Author Time Post
Reply The time has come to take the (D) off of PPP Polling. (Original post)
grantcart Nov 2012 OP
Dark n Stormy Knight Nov 2012 #1
Cicada Nov 2012 #2
grantcart Nov 2012 #6
Jim__ Nov 2012 #3
grantcart Nov 2012 #5
Jim__ Nov 2012 #7
grantcart Nov 2012 #10
Jim__ Nov 2012 #22
grantcart Nov 2012 #23
outsideworld Nov 2012 #4
DemocratSinceBirth Nov 2012 #8
bornskeptic Nov 2012 #9
grantcart Nov 2012 #11
bornskeptic Nov 2012 #12
Xyzse Nov 2012 #13
grantcart Nov 2012 #17
Xyzse Nov 2012 #19
former9thward Nov 2012 #14
outsideworld Nov 2012 #15
former9thward Nov 2012 #16
outsideworld Nov 2012 #24
former9thward Nov 2012 #27
grantcart Nov 2012 #18
former9thward Nov 2012 #21
PoliticalBiker Nov 2012 #20
cthulu2016 Nov 2012 #25
grantcart Nov 2012 #29
oswaldactedalone Nov 2012 #26
grantcart Nov 2012 #28
DCBob Nov 2012 #30

Response to grantcart (Original post)

Fri Nov 9, 2012, 02:39 AM

1. Thanks. Saw this on Lawrence O'Donnell last night and meant to find a copy. The asterisks,for those

wondering, indicate the 10 polls that owhile ten "overestimated Obama strength." The others did the same with Romney.

Your analysis sounds reasonable to me. I'm surprised about YouGov being so high on the accuracy list. I take their polls and their questions often strike me as having an R bias.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Fri Nov 9, 2012, 04:37 AM

2. PPP is hired by Dems but can still be accurate

PPP does have an economic interest in success by Dems. It's ok.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Cicada (Reply #2)

Fri Nov 9, 2012, 07:13 AM

6. You have the cart before the horse.


PPP is hired by Dems BECAUSE they are accurate, not because they promote Democratic memes. PPP was the first to predict the upset of Brown in 2010. Finding out that we were going to lose Kennedy's seat was the last thing that Dems wanted to hear or believe.

PPP has an economic interest in being the most accurate. That is why Dem's hire them. If the Republicans want to stay with ass kissers like Rasmussen and Graves, fine with me but if they have any sense they would hire PPP.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Fri Nov 9, 2012, 04:55 AM

3. When I follow your link, I see a page that ranks PPP as third most accurate.

The following list ranks the 28 organizations by the predictive accuracy of their final, national
pre-election estimates (as reported on pollster.com).
1. Ipsos/Reuters
2. YouGov
3. PPP (D)
3. Daily Kos/SEIU/PPP
4. Angus-Reid*
5. ABC/WP*
6. Pew Research*
6. Hartford Courant/UConn*
7. Purple Strategies
8. NBC/WSJ
8. CBS/NYT
8. YouGov/Economist
9. UPI/CVOTER
10. IBD/TIPP
11. Democracy Corps (D)*
12. CNN/ORC
12. Monmouth/SurveyUSA
12. Politico/GWU/Battleground
12. FOX News
12. Washington Times/JZ Analytics
12. Newsmax/JZ Analytics
12. American Research Group
12. Gravis Marketing
13. National Journal*
14. Rasmussen
14. Gallup
15. NPR
16. AP/GfK


Did the page change since you posted? Or, am I looking at the wrong page?

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Jim__ (Reply #3)

Fri Nov 9, 2012, 07:10 AM

5. I am baffled. I just hit the link again and got the same list

http://www.fordham.edu/images/academics/graduate_schools/gsas/elections_and_campaign_/poll_accuracy_2012_presidential_election_110712.pdf

—Initial Report, November 7, 2012—
Costas Panagopoulos, Ph.D.
Department of Political Science, Fordham University
For inquiries: cpanagopoulos@fordham.edu or (917) 405-9069
For all the derision directed toward pre-election polling, the final poll estimates were not
far off from the actual nationwide voteshares for the two candidates. On average, preelection
polls from 28 public polling organizations projected a Democratic advantage of
1.07 percentage points on Election Day, which is only about 0.63 percentage points away
from the current estimate of a 1.7-point Obama margin in the national popular vote
(Obama 50.1% versus Romney 48.4%).
Following the procedures proposed by Martin, Traugott and Kennedy (see Public Opinion
Quarterly, Fall 2006, pp. 342-369) to assess poll accuracy, I analyze poll estimates from
these 28 polling organizations. Most (18) polls overestimated Romney support, while ten
(10) overestimated Obama strength (indicated with a * below), but none of the 28 national
pre-election polls I examined had a significant partisan bias.
The following list ranks the 28 organizations by the accuracy of their final, national preelection
estimates (as reported on pollster.com).

1. PPP (D)*
1. Daily Kos/SEIU/PPP*
3. YouGov*
4. Ipsos/Reuters*
5. Purple Strategies

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Reply #5)

Fri Nov 9, 2012, 08:42 AM

7. It looks like that link takes me to an updated page.

http://www.fordham.edu/images/academics/graduate_schools/gsas/elections_and_campaign_/poll_accuracy_2012_presidential_election_110712.pdf

Note the UPDATED time:

Poll Accuracy in the 2012 Presidential Election
—Initial Report, November 7, 2012 (UPDATED 3:30PM)—
Costas Panagopoulos, Ph.D.
Department of Political Science, Fordham University
For inquiries: cpanagopoulos@fordham.edu or (917) 405-9069
For all the derision directed toward pre-election polling, the final poll estimates were not far
off from the actual nationwide voteshares for the two candidates. On average, pre-election
polls from 28 public polling organizations projected a Democratic advantage of 1.07 percentage
points on Election Day, which is only about 1.13 percentage points away from the current
estimate of a 2.2-point Obama margin in the national popular vote (Obama 50.3% versus
Romney 48.1%).
Following the procedures proposed by Martin, Traugott and Kennedy (see Public Opinion
Quarterly, Fall 2006, pp. 342-369) to assess predictive accuracy, I analyze poll estimates from
these 28 polling organizations. Most (22) polls overestimated Romney support, while six (6)
overestimated Obama strength (indicated with a * below), but none of the 28 national preelection
polls I examined had a significant partisan bias.
The following list ranks the 28 organizations by the predictive accuracy of their final, national
pre-election estimates (as reported on pollster.com).
1. Ipsos/Reuters
2. YouGov
3. PPP (D)
3. Daily Kos/SEIU/PPP
4. Angus-Reid*
5. ABC/WP*
6. Pew Research*
6. Hartford Courant/UConn*
7. Purple Strategies
8. NBC/WSJCostas Panagopoulos, Ph.D.
...


It still reflects PPP as an accurate poll - not a big deal - it may have to do with the page being retained in memory.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Jim__ (Reply #7)

Fri Nov 9, 2012, 10:42 AM

10. not a big deal but when I hit the link I get the other page, and that page is the one that is being

widely reported.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Reply #10)

Fri Nov 9, 2012, 04:06 PM

22. Here's a direct link to Fordham's report.

http://www.fordham.edu/Campus_Resources/eNewsroom/topstories_2590.asp

The reports on the internet that have PPP first were made before 3:30 on the 7th. Panagopoulos updated the report at 3:30.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Jim__ (Reply #22)

Fri Nov 9, 2012, 04:51 PM

23. you know that is very sloppy work



Its fine to make an edit but they should note it, especially when they know that it has been sent out into the internet.

Makes you wonder how thorough their analysis is if their presentation is sloppy as that.

Thanks for clarifying it.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Fri Nov 9, 2012, 05:19 AM

4. Always said that PPP ,was the most accurate pollster

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Fri Nov 9, 2012, 09:07 AM

8. In A Perfect World Republican And Democratic Pollsters Would Produce The Same Results

Social science research is social science research.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Fri Nov 9, 2012, 09:57 AM

9. When all the votes are counted, there will be a Democratic pollster on top of the list, but not PPP.

It will be Democracy Corps. Obama is currently ahead by more than 2.5% in the total, with a ton of votes still out which will favor him. He;ll end up ahead by about 3%.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to bornskeptic (Reply #9)

Fri Nov 9, 2012, 10:43 AM

11. No is talking about the largely irrelevent national polls but the hundreds of critical state polls


I am not even sure that PPP had a national poll, if they did I didn't even pay attention to them.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Reply #11)

Fri Nov 9, 2012, 11:44 AM

12. You didn't read the last line above the list of polls?

"The following list ranks the 28 organizations by the predictive accuracy of their final, national
pre-election estimates (as reported on pollster.com)."

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Fri Nov 9, 2012, 11:49 AM

13. NPR at 26?

That is pretty embarassing.
They got beat by Gravis... GRAVIS!
I feel like they fell.

I am enjoying some of their coverage again, now that the elction season is over, but geeze.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Xyzse (Reply #13)

Fri Nov 9, 2012, 12:27 PM

17. I don't take Gravis rating seriously. If you are going to run a con then you are going to select a

few polls to really inflate your numbers and keep most of them within the middle of the pack.

In fact if I was going to run a con with a new pollster I would start out favoring the opposite side and then slowly bring it around to the other side.

You may be interested in this thread

http://www.democraticunderground.com/10021774629

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Reply #17)

Fri Nov 9, 2012, 12:41 PM

19. Interesting...

I can agree with what has been shown, and I see how by inflating numbers during the campaign it could help perception and create that lead.

Then go for actual numbers near the end so that their polling doesn't look that bad in retrospect.

Either way, it still goes to show what I was saying; that NPR got out behind Gravis in regards to their polling, is an embarassment in my view for NPR.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Fri Nov 9, 2012, 12:01 PM

14. PPP did well on the national level but they were caught fudging figures in the MO Senate race.

They had Akin up by 1 after his rape comments when every other poll had him down by 9-12. They were changing their figures to keep him in the race. Akin even cited that poll as a reason to stay in. They changed their numbers after they were called on it. If that manipulation is acceptable to you from a pollster then fine. I would rather see accurate figures.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to former9thward (Reply #14)

Fri Nov 9, 2012, 12:11 PM

15. they did state , that most polled had probably not heard about the rape comments

Reply to this post

Back to top Alert abuse Link here Permalink


Response to outsideworld (Reply #15)

Fri Nov 9, 2012, 12:16 PM

16. Other polls released at the exact same time showed Akin down by 9-12.

That was just cover on their part.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to former9thward (Reply #16)

Sat Nov 10, 2012, 10:38 AM

24. Rasmussen poll was intentionally bad because they wanted to force akin out of the race

Reply to this post

Back to top Alert abuse Link here Permalink


Response to outsideworld (Reply #24)

Sat Nov 10, 2012, 11:31 AM

27. Mason Dixon had McCaskill up 9 at the same time.

Gravis had her up also. PPP was caught and changed their numbers a week later to conform to the other polls.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to former9thward (Reply #14)

Fri Nov 9, 2012, 12:38 PM

18. I am unaware of PPP doing any national polls



They are very well known for state polls.

If you work only the data then you will have outliers time to time.

But PPP follows the data. Most of the time it puts them on the top.

Sometimes they are their by themselves like when they called it for Brown in MA long before anyone else did.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Reply #18)

Fri Nov 9, 2012, 12:54 PM

21. From your link:

The following list ranks the 28 organizations by the predictive accuracy of their final, national
pre-election estimates (as reported on pollster.com).


Actually the most accurate is not even listed. RCP had Obama up 2.5% in their final average of polls.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Fri Nov 9, 2012, 12:44 PM

20. Accurate Polling...

is like facts to the right... when it is in thier favor, they are fine with it. When it's not, they make up their own.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Sat Nov 10, 2012, 11:00 AM

25. The D is not an indication of slant, it is what it is

PPP is a Democratic pollster.

Their clients are Democrats.

They editorialize on twitter. They ask prank questions to make republicans appear foolish.

They are a partisan pollster.

That does not mean their results are flawed. It means only what it means.

It is a necessary identification. If people chose to misread it as saying their results are suspect then that's what it is.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to cthulu2016 (Reply #25)

Sat Nov 10, 2012, 12:33 PM

29. That is basically my point


Gravis says that they are independent and get no letter. Neither does Rasmussen.


PPP may or may not be Democratic (and I accept what you say) but their results are objective.


The (D) or (P) should be based on actual leaning and not on other considerations because people reading it assume that if it has a D or P then its because its results are leaning not because of other factors.


Since the letters appear only to be a way of diminishing Democratic pollsters they should do away with all of the letter designations.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Sat Nov 10, 2012, 11:11 AM

26. While the Rand poll

was only a national poll and it was very unique in it's methodology, it came within a few tenths of nailing the general election results. I think they should feel good about their methodology and use it again in '16.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to oswaldactedalone (Reply #26)

Sat Nov 10, 2012, 12:28 PM

28. what was unique about their methodology

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Original post)

Sat Nov 10, 2012, 12:37 PM

30. Totaly agree. The perception that D pollsters are D leaning is in most cases totally wrong.

However on the flip side almost all the R pollsters are definitely R leaning. Hopefully the media has figured that out.

Reply to this post

Back to top Alert abuse Link here Permalink

Reply to this thread