the forest for the Iraqi trees

The Associated Press story from January this year breathlessly reports that:

“U.S. soldiers in Iraq are killing themselves at an unusually high rate, despite the work of special teams sent to help troops deal with combat stress.”

You have to scan a lot lower in the story to read the following:

“[T]he military has documented 21 suicides during 2003 among troops involved in the Iraq war. Eighteen of those were Army soldiers… That’s a suicide rate for soldiers in Iraq of about 13.5 per 100,000… In 2002, the Army reported an overall suicide rate of 11.1 per 100,000.

“The overall suicide rate nationwide during 2001 was 10.7 per 100,000, according to the federal Centers for Disease Control and Prevention.”

“An unusually high rate”? It’s only marginally worse than the suicide rate in the Army during peacetime, and indeed than the “civilian” suicide rate.

Now consider the fact that active military service overseas tends to be more stressful than normal life back in the United States. Also consider that US forces in Iraq are composed mostly of young males between the ages of 18 and 35 – the suicide rate for that demographic back in the US is almost 21 per 100,000, which is 7 per 100,000 more than for soldiers in Iraq (the actual figure I got for 2001, for “suicide injury deaths and rates per 100,000; all races, males, ages 20 to 34” on the National Centre for Injury Prevention and Control website calculator was 20.74). So a little bit of context to the story would have showed that in terms of suicide, Iraq is actually safer that Iowa.

But again, that would spoil a good story.
—the forest for the Iraqi trees  (Chrenkoff)

I’m blogging this becuase I’m often frustrated by how uncritically students accept statistics they read online. This analysis, written by somebody who makes no secret of his political opinion, makes an interesting case study.