The Plank

Lies, Damn Lies, And Ignorance Of Statistics


Matt Yglesias has graciously flagged his favorite part of the Gore backlash piece in today's New York Times. It's not bad. But I'm kind of partial to this graf:

Still, Dr. Hansen said, the former vice president's work may hold "imperfections" and "technical flaws." He pointed to hurricanes, an icon for Mr. Gore, who highlights the devastation of Hurricane Katrina and cites research suggesting that global warming will cause both storm frequency and deadliness to rise. Yet this past Atlantic season produced fewer hurricanes than forecasters predicted (five versus nine), and none that hit the United States [emphasis added].

I know almost nothing about global warming science, so that last line could be meaningful in a way I'm not seeing. But I do know a little bit (okay, a very little bit) about statistical inference. And everything I've been taught tells me that trying to draw inferences from a sample size of one is a pretty hopeless exercise. Which is to say, I have a hard time imagining how this past year's hurricane season matters one way or the other to any serious analysis of global warming. It reminds me of all the laughs conservatives had when a congressional hearing on global warming recently got cancelled on account of snow. Get it? It snowed one day this winter so global warming must be a myth...

Anyway, I'd expect that sort of thing from The Corner. I expect a little more from the Times. Maybe it's worth pointing out, though, that The Corner doesn't expect more from the Times, for whatever that's worth.

--Noam Scheiber

For more stories, like the New Republic on Facebook:

Loading Related Articles...
The Plank

More articles tagged as

Article Tools

Show all 22 comments

You must be a subscriber to post comments. Subscribe today.