ConWebBlog: The Weblog of ConWebWatch

your New Media watchdog

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

Friday, June 3, 2011
Fact-Challenged Corsi, O'Leary Team Up for Obama-Hating Book
Topic: WorldNetDaily

If there was anyone who believed Jerome Corsi was acting out of anything but hatred of President Obama when he wrote his dubious birther book, those illusions should be completely dismantled by Corsi's newest work.

Yes, less than two weeks after "Where's the Birth Certificate" was released, WorldNetDaily is publishing another book by Corsi: "How Obama Can Be Defeated in 2012: A Battle Plan Based on Political Statistical Realities," co-authored with Brad O'Leary.

WND is even promoting it as "a kind of sequel to" Corsi's birther book, and it quotes Corsi as saying, "This may be the last best hope for ensuring Obama is not back in Washington in 2013," so they're not even bothering to keep up any pretense that Corsi doesn't have an biased anti-Obama agenda.

The book supposedly contains "the most extensive polling of likely voters conducted by anyone," which supposedly proves that "Obama is very beatable in 2012 – if Republican candidates and strategists follow some simple cues":

Polling for this mammoth research project was commissioned by "The O'Leary Report" and conducted by IBOPE Zogby International. It surveyed 10,000 likely voters – a massive sample size that you simply will not find in other polls due the high cost and sophistication required to pull it off. In contrast, your typical New York Times poll, which serves as fodder for daily talk shows and opinion columnists, has a sample size of fewer than 1,000 people and frequently surveys adults as opposed to "likely voters."

As we've detailed, however, Zogby has a lengthy history of inaccurate polling, and the Zogby polling O'Leary has previously paid for is filled with slanted questions and misleading premises designed to provide the results he's paying Zogby to produce. There's no reason not to think this polling is any different.

WND's suggestion that O'Leary spent a large amount of money to obtain a sample size of 10,000 for his presumably skewed questions is likely not true. As Zogby itself states:

Zogby has amassed a database of respondents that numbers in the hundreds of thousands and is growing by the day. These people have agreed to take online surveys from time to time, for no compensation whatsoever. They never know when they will be invited to take a survey, nor do they know the subject. This database is constantly expanding, and Zogby technicians are constantly cleaning the database of obsolete entries and updating data points on respondents.

When a survey is initiated, a random sample is drawn from this pool of potential respondents, and Zogby sends them an invitation to participate in a survey via email, which includes a link that will take them to the survey on Zogby's own secure servers. The link expires after one use, which is just one of many security measures Zogby has in place to guarantee the veracity of the polling methodology.

One can probably assume that Zogby's method of gathering a database of "hundreds of thousands" predisposed to answer polling (as opposed to the time-honored method of a random sample from a cross-section) does not generate costs exponentially more than a typical sample size. And as analyst Nate Silver points out, Zogby's online method has generated results more inaccurate than even other interactive online polling.

Given that neither Corsi nor O'Leary are known for their scrupulous factual accuracy, this joint effort is probably twice as inaccurate than either man working separately.

Posted by Terry K. at 8:33 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

« June 2011 »
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

Bloggers' Rights at EFF
Support Bloggers' Rights!

News Media Blog Network

Add to Google