Friday, December 12, 2008

The most useless stastitic I heard today...

In an otherwise excellent lecture on animal abuse, the lecturer was talking about a study that compared hundreds of convictions for domestic abuse, looking at what predictors there might have been. She said:

"They found that convictions for animal abuse were a better predictor of future convictions for domestic assault than were convictions for murder, arson, and firearms crimes."

So, what's the big problem with that finding?

3 comments:

Dad said...

Dunno. Small sample? Murderers and arsonists are in prison? Football players are so busy shooting themselves they don't have time to shoot their wives?

Mom said...

Murderers and arsonists are in prison for a long time, which makes it somewhat difficult for them to beat up their wives. Animal abusers, on the other hand, probably got convicted of misdemeanors rather than felonies, served little jail time if any, and went home mad and looking for somebody to take it out on.

Did anybody point this out to the lecturer, Caleb?

Caleb said...

Nice work. The real point I was thinking about was the prison terms. Murder's probably not a good predictor of spousal abuse because murders go to jail for 30 years. It's hard to commit spousal abuse in prison.

No, no one pointed it out, but it was a minor point and she went by it pretty quickly. It stuck in my craw, though.