Cross-cultural surveys

I've noted this before, but, when faced with results like this, my first question is always: how the heck do we know people in different countries are even interpreting the questions in remotely the same way? For instance, how much smiling is "a lot"? Isn't that evaluation going to vary wildly between cultures?

But, at least if we have a number for it, we can feed it into other studies, and then average it! And, as I've noted, if we have five numbers, and we don't know what they mean, if we take their average, now we only have one number whose meaning is unknown. And that is surely a big improvement!


  1. Isn't that a big problem in survey data for the social sciences? I remember this being brought up a lot in my political science classes.

    1. Yes, I agree! The mania about "numbers" (which is why people turn to surveys instead of in-depth cultural studies) is a real problem.