MoltenThought Logo
"An appeaser is one who feeds a crocodile, hoping it will eat him last."
Sir Winston Churchill

3.13.2006

John Zogby, Partisan Hack Pollster

The latest Zogby fraud revealed:

The survey contains 24 questions. It was given secretly during late January and early February to an unknown number of American troops serving in Iraq, although we are told that 944 respondents were included. If all the guidelines for random sampling were met (they weren't), the reported margin of error would be plus or minus 3.3 percent.
The unforgivable flaw in Zogby's survey is the biased phrasing of its questions and answers. Two of the most provocative results are based on questions with no middle ground. It's like a multiple-choice test with no right answers.
For example, the widespread finding that three in four soldiers think the United States should withdraw from Iraq within a year has only one option for troops who think otherwise: stay indefinitely. This infamous question asks, "How long should U.S. troops stay in Iraq?" But the first three answers are not phrased in terms of staying, they are phrased "withdraw...," "withdraw..." and "withdraw... ." Where are the options for troops who think the United States should stay for "one to two years" or "two to five years"? Zogby omits such nuance. It's stay or go. Now or never.
The smart troops who perceived this false choice probably set the clipboard down and walked away at that point. That leaves us with a biased respondent pool.
Another question asks for a description of "your understanding of the U.S. mission in Iraq." Two choices describe the mission as clear, and four choices describe it as unclear.
More damning, John Zogby himself misrepresented the phrasing of one of the questions in an op-ed. This may seem like nitpicking, but if half a man's family say they want "chicken" for dinner, and he reports those votes as "nonvegetarian," he is not exactly being honest. In just this way, the poll asked the soldiers to rate seven different "reasons for the Iraq invasion." It is a question about prewar justification, not the postwar occupation. Yet Zogby described their answers as a description of "the U.S. mission." If that's the question he wanted to ask, he should have asked it that way. Polling is a science. Words matter.
The biggest question we should all be asking Zogby is not about the questions that were included, but about those that weren't. Nowhere in the survey results do we see assessments of the U.S. mission. Has it been a success or a failure? How so? Nowhere do we see questions about morale, about progress in killing terrorists, about the state of the insurgency, about the prospects for democracy and economic growth in Iraq. There are questions aplenty on napalm, interrogation, and (I'm not kidding) doubling the number of bombing missions.
Did Zogby dare to ask anything that might result in good news?


Of course not, anymore than he offered an unbiased view of exit poll data in 2004.

2 Comments:

Blogger linearthinker said...

MIM on Zogby:
"In addition to being a spokesman for the Wahabists, Zogby is regarded as a con man for hire, who manipulates results to suit his clients and further their stated agendas."

DJ Drummond at polipundit on Zogby:
"John Zogby may have a future in cable television, but not in professional polling. He’s become a media whore, and has crossed the line beyond which I do not believe he can return to credibility."

Drummond has fisked Zogby thorougly since the 2004 elections, including this recent Iraq "non-poll". Sample: "The poll simply failed to meet basic criteria for credibility, as is sadly common in Zogby polls these days."

The Iraq poll questionaire with demographics as provided by Zogby.

2:05 AM  
Blogger Teflon said...

Great links, LT--thank you.

What I find so disturbing about Zogby is that he invariably plays games with his data then tries to treat his methodology as though it were the Colonel's Secret Recipe.

Transparency in polling methodology is absolutely vital for assessing a poll's validity. It's the only way these results can ever be independently verified to any degree.

You know somebody's playing games when they claim to have some proprietary methodology that cannot be revealed to the uninitiated.

Unbiased surveys are very hard to craft and execute. Biased surveys are useless for anything but propaganda, and that only effective against the uninformed.

7:26 AM  

Post a Comment

<< Home