Pollster.com

Articles and Analysis

 

POLL: PPP (D) South Carolina Dem Primary


A new Public Policy Polling (D) automated survey in South Carolina (conducted 1/24) finds:

  • 595 Likely Democratic Primary Voters (± 4%)

    44 Obama
    24 Clinton
    19 Edwards
    14 undecided

Among a subsample of African-Americans (51% of the sample), Obama leads Clinton 67% to 13%.

 

Comments
Joshua D. Bradshaw:

Again just want to mention that fact that the polls coming out are so wildly different i do not think that their is a clear leader that can be derived from these polls and they are most likely untrustworthy. Now i must ask how is it possible to have so widely different poll number from the same time period.

____________________

RS:

I don't think the polls are completely untrustworthy - each poll is a sample of the population and thus has some uncertainty associated with it.
The PPP poll shows Obama 44%, Clinton 24%, Undecided 14%. With an error of +/-4%.
The Ebony/Jet poll says 37-27-22, error +/-2.9%.

So
(a) as Ciccina points out in the Ebony/Jet page, the variation in undecideds is large - that could be a factor of how hard the respondents are pushed (automated vs non-automated?)
(b) The numbers are within the polling uncertainties. Obama is 44 or 37, a difference of 7+/-5% (squaring the uncertainties on each number).

Though the difference seems somewhat significant compared to the uncertainty, combining (a) and (b) could explain the differences in the polling numbers.

Don't throw the baby out with the bathwater!

____________________

Adam:

One important point to consider is that a given individual's preference is not deterministic; in essence, it's a random process, with some associated probabilities of supporting a given candidate at a given moment. In other words, even if two different polls sample the same individuals, just from the fact the samples are taken at different moments can lead to wildly different results as these voters mull their options. It's important to pay attention to the numbers which state how likely the individual is to change their opinion; only if they're absolutely certain about their choice can you view their vote as a certainty in a given column.

That all being said, it's a game of expectations. You can get a good idea of Clinton's chances by putting a lower bound on Obama's support (which seems to be about 38%) and putting an upper bound on Clinton's (about 30%). From this, Clinton does have a chance if the undecideds break for her. But undecideds usually break for the challenger, and unlike in NH where she was the "challenger" given the utterly outrageous coverage of the press which put a nail in her coffin before people had even voted, she's the presumptive front-runner for the nomination here. Additionally, the editorials have ripped her quite a bit lately. Finally, she can't have a moment where she seems incredibly likable by tearing up, given her low-ball tactics.

In short, my thought is that Obama will receive at least 45%, carrying about half of the undecideds, and a lot of the other undecideds will break for Edwards, causing a battle for 2nd. I'll be in shock if Clinton somehow manages to win.

____________________

Richard Murray:

In biracial matchups "undecided" folks who go to the polls usually break strongly for a viable candidate of their own race. In the Democratic primary, with virtually no blacks participating, Senator Clinton likely benefitted from this pattern, as Barack Obama apparently got almost none of the self-declared undecided vote on primary election day.

South Carolina obviously presents a very different situation as many undecided voters are African American in this case, which augers well for the senator from Illinois. Plus, white undecideds have two options if they stick to their racial group, while blacks have only one. So, if the usual pattern holds, Obama should widen his lead as he gets a boost from the SC undecideds instead of the boot in New Hampshire.

____________________

Andrew:

By ignoring polling from states such as Massachussets, Connecticut, NJ and Arizona, Pollster.com is unintentionally being biased in favor of Barack Obama.

South Carolina this, South Carolina that. We have been saturated with SC polls even before Nevada voted. Which is fine as long as other state polls are covered as well.

There are polls from other states out there, as I earlier mentioned. And these are states that will vote soon. Please be fair.

____________________

tim:

Edwards doesn't know what he wants, and he's burned bridges with a number of people and it's starting to come back to bite him in the ass. Saw a brilliant vid that captures his essence in 2 min...

http://www.youtube.com/watch?v=K8PECQDBr2w

That's right - John isn't going to get the nom, and neither Barack (who may not hate John) nor Hillary (who surely does) is going give him another shot at the VP position.

____________________

fourth:

Andrew, are you serious? Surely not. The states that get the most polling are the ones that get attention here, and the states that are voting oh say, TOMORROW have a large number of polls coming out. Don't worry your pretty little head, the mean SC graph with the mean orange trend line of doom will soon disappear from the front page, and the focus will shift to the Feb. 5th states as it should.

____________________

fourth:

Yeah tim that sure is a "brilliant vid." Really captures his essence. I wonder whether it was a Bush or Clinton fan who put it together? I can't tell the difference anymore to be honest. Posting that serves no purpose here, and has no relation to the polls in general nor the poll in this thread. Go peddle your crap somewhere else.

____________________

Andrew:

Fourth, in your latest post, you gave no explanation as to why the attention given to polls from other states should be ZERO.

So you say that, because South Carolina votes tomorrow, then NOTHING can be said about polls in other states?
You are not very bright.

I say tell us what is going on in other states. America is not South Carolina, even if the election were to be held 5 minutes from now.

Unless Pollster.com's ISP is charging them by the paragraph, in which case they would have to save space by using the current approach.

____________________

Andrew:

...First of all I don't like to post back to back comments, but I forgot to note that Real Clear Politics, for example, posts EVERY single polls they stumble upon, regardless of whether it is from Alabama, Montana or any other state of the United States of America.

Pollster.com is much better than RCP, of course, because RCP is a biased right-wing outlet when it comes to commentary, but as complete archive for polls, this issue could, IMO, be tweaked here to include more states.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR