From the Department of Ambiguity

Surveys should clarify, not further muddy, the picture. Pay attention to the reporting when results from this survey come in.


The military is, according to CNN, asking soldiers these questions about the possible repeal of Don't Ask, Don't Tell.

-Would a repeal of "Don't ask, don't tell" prompt a soldier to reconsider serving in the military?
-Would serving under an openly gay commander adversely affect morale?
-How would troops feel about sharing a bathroom or open-air shower with an openly gay comrade in a war zone?

The first question could result in ambiguous results. Presumably, a gay soldier might reconsider serving precisely because DADT has been repealed--in other words, he or she would feel safe remaining in service. But watch for interpretation that assumes that anyone who says he'd reconsider service means that the respondent is opposed to repeal.

 On the second, "openly gay." Well, let's turn that around. If someone is obnoxious about his sexuality, hetero or gay, that could be enough to affect morale.

 On the third, why limit it the question to the war zone? Soldiers have little or no privacy from day one of basic training.

 And why, honestly, are we asking soldiers this? Were soldiers surveyed before President Truman ordered the integration of the military? Don't think so. I wish someone would ask questions about this.  And dear gay friends, I respect your fears about soldiers outing themselves, but if you take yourselves out of the process, the results will be skewed.

Psychic Polls Alert

The Hill doesn't know how to read polls

As someone noted, you can't get good poll results from something that hasn't happened. Trying to guess the effect of a Senate vote on the health care bill when it has't occurred yet is foolish--why even report on them?

On Poll Bias

Nate Silver has a good assessment of how the order of questions can affect poll results. Question Order May Bias Health Care Polling

Every copy editor ought to be reading this guy.

Watch Those Polls

Sure, the presidential campaign is over for now but the polls haven't gone away, particularly the kind that Nate Silver at FiveThirtyEight identifies as a kind of push poll through skewed questions.

Numbers Guy faults it too. Zogby’s Misleading Poll of Obama Voters

During a campaign, pollsters can build credibility by forecasting election results accurately. Afterward, they can build revenue by using that credibility to attract private clients. These private surveys often have an agenda, and their numbers can’t be tested against an objective standard, such as votes. Such surveys can test pollsters’ standards of conduct.

Zogby International recently conducted a survey for a critic of president-elect Barack Obama and then, together with the sponsor, interpreted the numbers from the survey in a misleading fashion.

Then there was this little harmless-sounding poll the AP reported on Saturday, showing us once again how ignorant Americans are about their own political system. I happen to agree with that assessment, so was reading through the story about how we're all going to hell in a handbasket but found some strange writing. First, the story cited multiple problems with answers but didn't give the questions. Then suddenly, it dropped in this loaded gem:

The question that received the fewest correct responses, just 16 percent, tested respondents' basic understanding of economic principles, asking why "free markets typically secure more economic prosperity than government's centralized planning?"

Doesn't the wording strike you as odd? So I looked up the outfit that ran the poll, a group called the Intercollegiate Studies Institute, and noticed the names of several conservatives rotating across the top of the page, including Whittaker Chambers, William Buckley and others.

The group, of course, is entitled to push any idea that it wants. But along with the loaded attempt of the Obama poll, cited above, makes me, once again, say polls should be reported extremely carefully and the source--and its motivation--spelled out.

And on a side note, is it me or is it especially tacky for a cable movie channel to be showing that fantasy film, "JFK" today?

Those Polls

The Numbers Guy warns against poll averaging. Two companies

are mashing up surveys from various sources this election year to produce composite numbers meant to smooth out aberrant results. Their methods are criticized by statisticians, but their numbers are embraced by news organizations eager for a way to make sense of conflicting polls.