Pollster.com

Articles and Analysis

 

Judicial Watch Poll: Zogby Responds


Zogby International has posted a response to the criticism last Friday from various sources, including yours truly, of a poll they conducted for Judicial Watch. The argument should sound familiar:

Zogby International stands by its recent telephone poll conducted for Judicial Watch, a conservative interest group based in Washington. The survey probed the thinking of likely voting Americans about New York Sen. Hillary Clinton, and while some have charged that Zogby's questions were biased to solicit responses that would put Senator Clinton in a bad light, the results of the survey belie that charge. The responses to the Zogby survey were quite similar to the responses given to similar questions asked in an ABC News/Washington Post survey of May, 2006, and a CNN/USA Today/Gallup poll of August, 2005.

Zogby International maintains that its wording of these questions was not biased, largely because the topic - the trustworthiness of Ms. Clinton - has been at the center of exhaustive public discourse over the past 15 years and is one about which most American voters had drawn conclusions long before the Zogby survey. The similar results of the other two surveys, cited above, emphasize this fact.

Having made a similar argument using precisely the same survey results, I have to agree that most Americans have drawn conclusions with respect to Senator Clinton and her trustworthiness. But if Mr. Zogby believed so strongly that Americans had already made up their minds about Senator Clinton, why did he begin his question by reminding respondents that "some people believe that the Bill Clinton administration was corrupt?" Why not just ask, without the preamble, "how concerned are you that there will be high levels of corruption in the White House if Hillary Clinton is elected President in 2008?"

That preamble is the main reason critics like me concluded that Zogby's questions were "biased to solicit a response." Zogby's odd defense makes that intent even more obvious.

Why? Consider that pollsters frequently provide this sort of factual preamble when asking about an issue about which many respondents may be unfamiliar. One such example, noted by a commenter who disagreed with my criticism of the Zogby survey, is a question asked on a recent Newsweek survey about the recent controversy involving the Walter Reed Army Medical Center:

Since news reports last month about neglect and poor health care for military personnel returning from Iraq and Afghanistan at the Walter Reed Army Medical Center, do you think the Bush Administration has done a good job or a bad job of dealing with this situation?

Although the author of that question probably assumed that respondents would distinguish between the problems at Walter Reed and the way the Bush administration dealt with those problems, the commenter is right to see this question as a bit leading. Doesn't the existence of "neglect and poor health care" imply that a "bad job" has been done?

It is not hard to find similar examples. Media pollsters often use such preambles when asking about issues of great interest to political insiders that are nonetheless unfamiliar to most Americans, and accusations of bias often follow. Consider the recent controversies over questions on Terry Schaivo, the NSA's monitoring of domestic phone records or the Senate filibuster.

Whatever you conclude about the presence existence of bias (subconscious or otherwise) in those examples, all had one thing in common: Most Americans were unfamiliar with the facts involved. The pollsters included an introductory preamble in the hope of providing respondents with just enough background information to help answer the question.

But in defending his questions, Zogby provides a totally different explanation. The "trustworthiness" of Senator Clinton, he tells us, "has been at the center of exhaustive public discourse," something about which "most American voters had drawn conclusions long before the Zogby survey." Well if that is so, why did respondents need to be reminded that "some people believe that the Bill Clinton administration was corrupt?" What possible purpose does that preamble serve, except the hope that it would lead respondents to a desired answer?

 

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR