ConWebBlog: The Weblog of ConWebWatch

your New Media watchdog

ConWebWatch: home | archive/search | about | primer | shop

Saturday, July 11, 2009
O'Leary Baselessly Attacks NYT Poll, Puffs His Own
Topic: WorldNetDaily

Brad O'Leary uses his July 10 WorldNetDaily column to attack a New York Times poll he didn't like -- and, of course, to promote his own skewed polls.

O'Leary asserts that the Times"rigged the poll much like the Ayatollah rigged the Iranian election," claiming that it "heavily weighted the poll in favor of Obama, showing the gap between his personal approval rating and the public's approval of his initiatives to be much smaller than it actually is." O'Leary also complained that the poll "surveyed twice as many Obama voters than McCain voters, as well as a significant number of non-voters. ... Sixteen percent of those surveyed by the Times are not even registered to vote."

By comparison, O'Leary asserted, "the O'Leary/Zogby poll used an honest and accurate sampling method by only surveying Americans who voted in the 2008 presidential election, and weighting that sample to reflect the actual outcome of the election."

O'Leary's attack on the Times poll -- portraying it as "rigged" and "portrait of a mythical America that doesn't exist," while puffing up his own as "honest and accurate" -- is without substance. O'Leary doesn't explain why non-voters are somehow less qualifed to offer opinions on presidential policies -- last time we checked, they were American citizens and subject to all the same regulations as voting Americans.

The Times' use of a different methodology than O'Leary doesn't make its results any less valid. As Janet Elder, the Times' editor for news surveys and election analysis, said in a July 24 CNS article, "Although some polling organizations do, The New York Times/CBS News poll does not weight by party ID. ... We weight by characteristics that are known from census data." O'Leary took the cop-out that "the Times failed to disclose the make-up of its sample in the article it published that detailed the poll's results" -- even though it could have been found by simple Googling -- thus avoiding having to respond to the Times poll's methodology.

Further, as Slate points out, it's highly likely that the large disparity between declared Obama voters and declared McCain voters in the Times poll is because people aren't telling the truth to pollsters because they want to be on the side of the winning candidate (and don't want to be associated with a loser like McCain):

What gives? Are people really lying about having voted for Obama?

Yes, they are. It's common for more people to claim they voted for a president than actually did. In the 1930s, George Gallup found that Franklin Delano Roosevelt was more popular in post-election polls than he was on Election Day. The same was true after the 2000 election, in which George W. Bush lost the popular vote. By 2004, polls showed Bush having won in a landslide.


The main explanation for the gap, say pollsters, is people who didn't vote at all saying they did. These people tend to say they picked the winning candidate. Just look at the Times and Journal polls, where about 80 percent of respondents said they voted in the 2008 election. In fact, turnout was about 61 percent. (A 20 percent gap is pretty standard.) Pollsters attribute the disparity to the social discomfort of having to admit, even to a stranger on the phone, that you didn't vote. Exacerbating the discomfort is the fact that the question "Who did you vote for?" usually comes at the end of a survey—after you've just spent 30 minutes telling the pollster what you think of Obama. What are you going to do, admit you never voted?

Another reason is forgetfulness. If you've read this far, you're probably pretty interested in politics, and maybe you have indelible memories of Election Day 2008 seared into your hippocampus for all time. But most Americans don't pay close attention to politics. Plus, people do a poor job of reporting past behaviors. Studies show that patients have a hard time remembering when they visited the doctor, let alone what their doctor told them. Same with voting. Say you normally vote but can't quite remember whether you voted in the most recent election. You might well say you did. And because you like how Obama's doing so far, you figure you probably did vote for him.

Then there's the group of McCain voters that either regrets their pick or would rather not admit it to a pollster. They might feign forgetfulness, which would account for the 7 percent of respondents who say they voted for "someone else" or won't say for whom. Or they might just say they picked Obama. But outright dishonesty probably accounts for little of the gap.

O'Leary then offered "a comparison of both the New York Times' and The O'Leary Report's findings on similarly asked poll questions." But as O'Leary surely knows, "similarly asked" does not mean the same thing as exactly asked; slight changes in wording can produce different responses. As we've detailed, the Zogby polls O'Leary pays for include questions tweaked to obtain the response he wants.

O'Leary largely fails to provide the questions specifically asked by both polls, and the one example he provides shows O'Leary's bias:

Both the Times and O'Leary/Zogby asked Americans similar questions regarding how large a role government should play in society.

The Times asked: "Which comes closer to your view: Government should do more to solve national problems, or Government is doing too many things better left to businesses and individuals?"

In the Times' manipulated sample, 34 percent think the government should do more, 56 percent think the government is doing too much, and 10 percent are unsure.

O'Leary/Zogby asked Americans whether they prefer "a system in which the public or the state have ownership and administration of the means of production and distribution of goods," or "a system in which wealth, and the means of producing wealth are privately owned and controlled rather than state owned."

In O'Leary/Zogby's accurate sample, 70 percent prefer a private system, 17 percent prefer a state-run system, and 13 percent are unsure.

By using phrases like "means of production" and "state owned" -- phrases long used in reference to socialism and communism -- O'Leary is using language that is clearly designed to strike a negative subliminal response, predisposing respondents to reject that option. Thus, it can be argued that O'Leary's sample is not only not "accurate," it's just as "manipulated" as the Times poll he denigrates.

It's also worth noting that for as much as O'Leary complains about the Times' alleged failure to "disclose the full details behind its sampling method," at no point in his article does O'Leary offer a link to his own Zogby poll so we can examine the methodology and specific questions he asked for ourselves.

Posted by Terry K. at 11:03 AM EDT
Updated: Saturday, July 11, 2009 11:07 AM EDT

Newer | Latest | Older

Bookmark and Share

Get the WorldNetDaily Lies sticker!

Find more neat stuff at the ConWebWatch store!

Buy through this Amazon link and support ConWebWatch!

Support This Site

« July 2009 »
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31

Bloggers' Rights at EFF
Support Bloggers' Rights!

News Media Blog Network

Add to Google