Friday, April 20, 2007

Let's play 'Month of Tragedies'

Stay tuned: Editing exercise at end!

One of the little-known benefits of being a copy editor -- leaving aside the massive salaries, the luxurious working conditions, and the hordes of grammar groupies -- is the unlimited Emperor's New Clothes license issued when you take the nightside's shilling. With it, you can stand up at any point when the parade is going by and proclaim that the emperor has no clothes.

Which is what somebody should have done (or done louder; I wasn't there) with the story in question. The experts are right. It's coincidence. April is a "month of tragedies" because every month is a month of tragedies (by which the writer apparently means "unrelated killings," not disasters, interstate violence or plays in which a central flaw leads to grisly doom for the main character).

Let's look at the story:

In April 1995, two men executed what was then the worst terrorist act in American history by bombing a federal building in Oklahoma City. (So American history starts -- when, late 1983? Just wondering.)

On an April day four years later, Columbine High School became the site of the nation's deadliest school shooting. (Though not, by a long chalk, the deadliest attack on a school.)

Now, almost exactly eight years later, the mass murder at Virginia Tech has set a new, horrible standard: A single gunman killed 32 victims.

Is there's something about the month of April? (I don't know. Is there's?)

Criminologists and psychiatrists say it's unlikely. The timing of these incidents are coincidences, not a trend.

Here's the part where you get to stand up and say "He's naked!" Sorry about the phone calls the writer has already made, but at this point it's clear that the story shouldn't run. Actually, it should have been clear after the first couple of phone calls. But while we had the psychiatrists on the line, it might have been nice to ask them about cognitive biases, because what we have here is a blind spot in how journalists evaluate evidence.

That's not a moral failing or anything. We all put a heuristic thumb on the judgment scales every now and then. The problem in this case is that someone decided to turn faulty reasoning into a 1A presence, and that's a bad idea.

Mass mayhem in April is a very "available" construct right now. Columbine has been talked up a lot, and when you add in Oklahoma City (even though seasonal affect has exactly the square root of zero to do with political violence), you have what looks like a Trend. But the emperor's naked. You can find trends every bit as good almost anywhere, as long as you trust the data instead of your lyin' forebrain.

Today's editing exercise, then, is to play the "month of tragedies" game yourself. Pick a month and find at least three acts of substate mass killing (for master-level play, stay within U.S. borders and don't count political violence and terrorism):

August: Month of tragedies?
Edmond, Okla., 1986 (Aug. 20): Post office shooting, 14 dead
Austin, Texas, 1966 (Aug. 1): Campus sniper, 14 dead
Fayetteville, N.C., 1993 (Aug. 6): Restaurant shooting, 4 dead (BONUS: Name the Clint Eastwood movie the gunman had been watching before the attack).

Who's got one?

2 Comments:

Anonymous Anonymous said...

Sunday: Day of bloody tragedies
-March 7, 1965: Violence during civil rights marches in Alabama
-Nov. 13, 1887: Violence in London during demonstration against coercion in Ireland
-Jan. 30, 1972: Twenty-six protesters shot in Northern Ireland

The list goes on and on and on ...
http://en.wikipedia.org/wiki/Bloody_sunday
See?

And, Fev, thanks. I laughed out loud when you asked, "Is there's?"

6:04 PM, April 20, 2007  
Anonymous Anonymous said...

I love how often I'm seeing the new generation of "trend" story -- the "not-a-trend" story!

I saw this Dispatch hedline the other day and didn't even both glancing at it, as I knew it would be something like this.

The thing is, it seems there could be some interesting angles in relation to this, if someone wanted to take the time and effort to scope them out. From being in a liberal-arts college with a lot of stress and many students emotionally on edge, I'd be curious whether there is data on any correlation between time of the school year (i.e., a month after freshmen arrive and might feel overwhelmed, or under the stress at year's end) and dangerous actions. Again, though, just about any time of the school year could have factors that could increase anxiety... And a lack of correlation would seem a reason to not run a story, rather than run a story on... well... nothing. Still, finding out any of that would take some more serious research than some papers seem to want to spend...

9:24 PM, April 21, 2007  

Post a Comment

<< Home