Troops fear they're being stretched too thin, poll says

[Barrett808]

Troops fear they're being stretched too thin, poll says
Military Times
Dec. 29, 2003 12:00 AM

WASHINGTON - Despite a year of constant combat casualties and grinding overseas tours, men and women in uniform back President Bush and his policies in Iraq, according to a Military Times poll.

But their support for Iraq policy isn't much higher than that found in the general population. And while a number of morale indicators are positive, the poll found overwhelming sentiment that the demands of the war on terrorism have stretched U.S. troops so thin that their effectiveness has been damaged.

The findings are part of a Military Times poll of 933 active-duty military members who subscribe to the Army Times, Navy Times, Air Force Times and Marine Corps Times. The subscribers were randomly surveyed by mail in late November and early December. That group, though it tends to overrepresent those making the military a career, is perhaps the most representative sample possible because of the inherent challenges in polling members of the military, according to polling experts and military sociologists.

The poll carries a margin of error of plus or minus 3.3 percentage points. Among its findings:

• Even before the capture of Saddam Hussein, which came after the survey, the poll found 56 percent of those in the Military Times poll approved of Bush's handling of Iraq.

Still, those numbers are not much higher than support in the United States as a whole.

• Seventy-seven percent of those polled agreed with the statement that the military is stretched too thin to be effective; only 14 percent disagreed.

http://www.azcentral.com/news/articles/1229militarypoll29.html

[JeebusH]

1. ...the poll found 56 percent of those in the Military Times poll approved

who wants to bet that most of the 56 percent AREN'T in Iraq right now

[tabasco]

6. Really.

Let's take a poll of Army grunts.

[arcane1]

2. 56%, huh? You would expect a higher number from the article

first sentence:

"Despite a year of constant combat casualties and grinding overseas tours, men and women in uniform back President Bush and his policies in Iraq, according to a Military Times poll"


uh, actually nearly HALF do NOT support Bush and his policies. Why isn't THAT news?

:shrug:

[Sagan]

3. a subscriber poll?

Doesn't that have some kind of bias to begin with?

[Journeyman]

4. I seriously question that a subscriber poll. . .

of a group that "tends to overrepresent those making the military a career," would have a 3.3 percent margin of error. That's a MOE in line with what would be expected from a random blind sample, not one incurred in a poll of the choir.

[daleo]

5. Statistical margin of error of a sample of 933 is about 3.2%

But, in this case all that means is that if you sample the entire population of subscribers the true level of support is likely to be within 3% of the reported figure, using a 95% confidence interval.

But, this is a biased sample, so it says nothing about the population of military people in general. It's like using subscribers to playboy magazine to construct a sample frame, then using that result to comment on the general population's attitude toward censorship.

I can't understand their claim about this being the best sample frame they could come up with - more likely it is one that they hoped would be most supportive of the outcome they wished to report.