Saturday, April 12, 2014

Dartboard geography

With American journalism now under the rule of the wonks, the above map is getting a lot of attention. It was featured on a post on The Monkey Cage blog at the Washington Post and later written up on the Klein-Yglesias supergroup vehicle Vox. The dots represent the responses of each of around 2000 survey participants to a graphical challenge to pick out Ukraine on a map, and the roars of laughter come from wonks everywhere with the accompanying finding that the people worse at picking out Ukraine on the map were more likely to favour US military intervention in it.

Fine. Leave aside the fact that there's little evidence that the people who can't pick out foreign countries on maps are the ones who are influential. Billy Kristol could pick Iraq out on a map in 2003. What's remarkable about the map responses is how widely distributed they are once you take out the eastern Europe and Central Asia clusters. Just about every African country got a guess, most of Latin America, a few US states, lots of Canada including its Arctic territories, quite a few who thought it was in Greenland, a fascinating straight line of dots down the middle of Russia.

In short, there's circumstantial evidence that many survey respondents did not meaningfully participate in this survey at all. So the conclusion is not that people less likely to know where Ukraine is more likely to favour military intervention. It's that people who didn't put much thought into this survey were more likely to favor such intervention. This sounds like a ripe scenario  for presentational anomalies.