Kieran Healy

It’s called Sampling on the Dependent Variable

“Why are there so many personally obnoxious people on the Left?” asks Nick Denton. In his next sentence, he explains to himself why his question is mistaken. (“I happen to spend most of my time in cities which tilt Left … so the obnoxious people I meet are statistically more likely to be left-wing.”) But he presses on regardless to the inevitable single data point. A couple of months ago, using the same methodology, Dan Drezner wondered why people on the left were more sensitive to insults than those on the right. So with that data in mind, we can now conclude that left-wingers are both more sensitive and more obnoxious than right-wingers.

Coming up soon, a study that attempts to explain why so many of the closets I open are full of my clothes.

Nature loves a tautology.

Self-fulfilling prophecies also flourish.

Wow. And I thought I had perfected the extrapolation from a single data point method.

I don’t believe your unsubstantiated claim that many of the closets you open are full of your clothes. It seems extremely improbable, because I open closets almost every day and none of them contain anything belonging to you.

Nick Denton talks about those who are ‘personally obnoxious’? Pot, meet kettle.