Pentagon sex assault survey raises doubts of accuracy

Article here. Excerpt:

'The Pentagon is mindful of the survey’s critics. Rand, in a sense, will be double-checking the accuracy of previous polls.

“The Defense Department contracted with Rand to have an independent, outside organization conduct the biennial prevalence survey,” said Army Maj. James Brindle, a Pentagon spokesman. “Rand’s survey methodology will allow the department to accurately compare data among all previous surveys so that we can seamlessly assess progress the department is making in preventing and responding to the crime of sexual assault.”

Critics do not point only to a supposed skewed sample. They say the Pentagon’s results do not mesh with society as a whole.

The Bureau of Justice Statistics reported a sex assault rate of just four-tenths of 1 percent for women age 18-34, a range that roughly matches the military population. The Bureau’s polling in 2010 relied not on Internet responses but on phone calls and personal interviews conducted by the Census Bureau.

Based on those results, the military has a rate of sexual assault against women that is 12 times higher compared with civilians. Overall, the Bureau of Justice Statistics reports that sexual assault rates have been falling in the U.S. over the past 20 years.

Critics also cite the Pentagon survey’s wide swings. In 2006, its poll found 34,000 total sexual assault victims; in 2010, the number dropped significantly to 19,000; and then in 2012, it went back up, to 26,000 — 12,000 of whom were women.'