Saturday, January 28, 2006

NYT: Op-ed News shows NYT how to report on polls

The / Zogby People's Poll is the first poll run by a progressive organization and possibly the first poll run by a media website, certainly by a progressive media website.
Zogby International conducted interviews of 850 likely Pennsylvania voters online on January 26th and 27th.
The poll consisted of 22 questions related to the PA race for US Senate and a total of over 70 questions, including 27 issue-related questions on impeachment, Bush's NSA spying and hearings, electronic voting, war in Iraq, withdrawal from Iraq, nationalized health care, privatization of voter registration and election technology and more. The PA Senate Race: Santorum vs. ?
The poll is the first to ask about all five PA US senate race candidates and provide positions each holds. When people find out about candidate positions on key issues, Casey plummets from a twelve point lead to a dead heat, with a non-significant two point lead. Santorum loses when matched with any of the democratic candidates. Both of the self identified progressive democratic candidates draw higher percentages than Casey, with Pennacchio having the largest percentage of votes against Santorum, at a not quite significant 4.4 points higher than Casey. Casey, at non-significant levels, actually gets MORE votes from Republicans once they find out more about him and Santorum.
Casey refused to respond, so we constructed his positions from media and speeches.'s conclusion on this race-- Santorum wants Casey as his opponent because he wins the demographic game. Casey loses massively in some categories, when voters find out about Casey, which Santorum will sure insure.
For example, Casy's support among 18-24 year olds drops from 63% to 40%, with Protestants, it drops from 47.3% to 30%, with liberals, from 95.4% to 68%, with moderates, from 64% to 53%, but Casey actually gains support from conservates, going from 3.9% to 5%, a non-significant, but interesting finding.
The poll first asked panelists how they would vote, matching up all five candidates (three Democratic, two Republican) without any information provided to them, then presented them with information on the position the candidates take on a number of issues. All candidates were invited at the same time to participate and given three days to respond. They were told that if they did not respond, then the issue positions would be constructed based on website, media and other statements the candidates made.
All candidates responded except Bob Casey, jr., who replied that he would not participate in the poll. made one last request by phone and email after the deadline, offering the Casey campaign an opportunity to participate and they stated they were not participating. That same evening, Casey was also a no-show at debate sponsored by the league of women voters in Harrisburg PA.

The above is from Op-Ed News' "OpEdNews.Com/ Zogby People's Poll." You didn't read it in the New York Times. You can tell that because instead of pat answers, you're told if something is significant or not. The sample size of the poll (of Pennsylvania residents) is roughly the same sample size that the Times/CBS polls use to convey the mood of the nation.

This is one poll. If subsequent polls bear out the same results, then you have a trend. In terms of the subject of the poll, Casey Junior, you also have history and the history (as noted in The Third Estate Sunday Review's "Robert Casey Junior Doing Pop Proud" on March 6, 2005) is that Casey Junior is someone that voters turn against as they learn more of him. He polls really well, early on. Then he peters out. As we wondered at the time (I helped with the piece):

Junior and Santorum both reject public schools for their children. They both are anti-choice. Does someone have some photos of Santorum engaging in man-on-dog sex? If not, why are they [soft-minded Dems] so sure that a carbon copy of Santorum can trump the original?
[. . .]
This is one of the most idiotic decisions soft-minded Dems have made in recent days (a tough call, granted). "We're going to run someone just like Santorum and we'll win!" Say what you want about Santorum, but he has a superficial physical attractiveness. (Janeane Garofalo has compared his looks to those of a gay porn star.) Is Junior with his receding hairline and near uni-brow really going to be able to stand on stage opposite him and look "Senatorial?" We're puzzled [by] that notion as well.

And noting Casey Junior's pattern of losing, we concluded with this:

Considering how easily most incumbents win re-election, soft minded Dems (who apparently fear the democratic process) better hope they have something along the lines of Man-on-Dog photos of Santorum. Otherwise, Junior's about to get spanked again.
[. . .]
We'd love to see Santorum out of the Senate but on election night, if Junior's standing in the corner, rubbing his heinie and bawling his eyes out, don't expect us to shed any tears. Though it's doubtful Junior will learn his lesson, maybe soft minded Dems will.

A member in Pennsylvania e-mailed that (as well as the next highlight) and wondered if we could highlight it "even though you hate polls." I really hate most summaries of polls where journalists fudge the facts as they reach sweeping conclusions that are "based on a true story" and about as close to reality as any real bad TV movie. Nagourney & Elder and their ilk should add "inspired by a poll" to their bylines.

Op-Ed News isn't presenting them as the last word, fudging the facts, or attempting to create a sweeping narrative from raw data. And, again, the sample size is the equivalent of what the Times/CBS polls use to represent the entire nation, only this poll focused on one state. Op-Ed News is an independent outlet and deserves applause and recognition for entering the polling practice. It's been left to many on the right, center-right and center for too long. So if anyone sees a poll they've commissioned (or anything else they provide, op-eds, what have you) always feel free to highlight it in an e-mail. We support indymedia here and that alone would have gotten it highlighted. The fact that indymedia's done a better job of reporting on a poll than the paper of record does is only all the more reason to highlight them.

This was the second highlight, here Rob Kall focuses on one question in the poll with "85% of Democrats are more likely to vote for a candidate who supports impeachment:"

The poll of 850 Pennsylvania panelists via internet interview by the Zogby organization was run on January 26th and 27th.
The exact question was:
34. How likely would you be to vote for a candidate who supports having impeachment proceedings against President Bush -- very likely, somewhat likely, or not likely?
1. Very likely
2. Somewhat likely
3. Not likely
4. Not sure
In the poll, 85% of Democrats reported they would be more likely to vote for a candidate who supports impeachment.
49.3% of independents would be more likely to support a candidate who Supports impeachment, while 40.6% of independents would be less likely to vote for a candidate who supports impeachment. 90.4% of Republicans ae less likely to support a candidate who supports impeachment.

There's more at Op-Ed News (much more) so feel free to visit their site.

How do you lose a long post? Pretty easily. (That's not a slam to the friend who's taking dictation of this entry.) This is going up and the Times commentary will be redone shortly. It was supposed to appear with this entry and it got lost when my friend saved to draft.

The e-mail address for this site is