Some time back I blogged about research in gender and political knowledge. The short version is this: studies repeatedly find guys do better than gals on tests of public affairs knowledge. One explanation? Men answer questions even when they don't know the answer and sometimes get it right, while women do not.
Here's another theory tested. Unfortunately I can only point to an abstract, not the full text, but the upshot is you can improve how women do when you ask questions that include female politicians.
This makes perfect sense.
One of my own studies explored with conservative Christians do so poorly on tests of political knowledge. In part it's demographics, but when you ask religious questions about candidates (their affiliation) versus what state they are from, the conservative Christians do not fare so poorly. The upshot? Ask people what they know, in a certain domain, and they do better than previous research suggests.
This has a lot to do with cognitive accessibility and all that rot. Fun stuff if you're into research, but translated it just means we have certain things that are important to us and when questions are posed in such a way as to tap those interests (versus those of scholars and news junkies) then we do okay. Better. Not great, mind you, but better.