Wednesday, February 21, 2007

Flawed Iraq Poll Being Flogged by Noise Machine

Apparently the NY Post ran a front today that a new poll, conducted by a national Republican polling firm, shows much greater support for, well, hanging around Iraq and hoping for the best, when you really look at the results. Here's how the Post put the best gloss on it:

    The poll found that 57 percent of Americans supported "finishing the job in Iraq" - keeping U.S. troops there until the Iraqis can provide security on their own. Forty-one percent disagreed.

    By 53 percent to 43 percent they also believe victory in Iraq over the insurgents is still possible....

    Only 25 percent of those surveyed agreed with the statement, "I don't really care what happens in Iraq after the U.S. leaves, I just want the troops brought home." Seventy-four percent disagreed.
Despite the breathless reporting of the poll results on various Noise Machine nodes, I find myself underwhelmed by that description. But wait, there's less.

Some liberal blogs launched some from-the-hip criticisms. Then TPM went over the questions with a Republican pollster of their own. His opinion: the poll is hopelessly biased. For you polling geeks (yes, Red, talking to you) the details are a must read, not only to debunk this particular poll, but as a window into the science of drafting good questions.

2 comments:

Jason Haas said...

oh, sweet mercy, that poll is junk, or more to the point, a POS.

Scott Piepho said...

Ah, yes. It's a real problem calling your firm Public Opinion Strategies -- people refer to the POS poll.