Thursday, July 11, 2013

10-Nation Study: Women Less Politically Knowledgeable

So I'm having sleepless night and this tweet comes across my screen at 3:14 a.m.:
Let's look closer at the research they're tweeting about. Basically the article about research says what's in the tweet above, but thankfully includes some caveats, such as:
Academic studies have previously found that women have higher levels of risk aversion and so are afraid of being wrong. When faced with multiple-choice questions, women are more likely to give a 'don't know' response than men. Others argue that the questions used to gauge political knowledge tend not to be gender-neutral and that women's political knowledge is more concerned with a personal experience of local politics and government programmes relating to daily life.
Simply put, men are more likely to guess on political knowledge tests (but not ask for directions). Lemme explain how this skews results. We generally score such tests as "1" if you answer correctly, "0" otherwise. That means being wrong or saying "I don't know" get scored the same. Men guess more, they're gonna get some of them right (and some wrong). They get extra points this way, by guessing. Studies that control for this tend to find this "guessing" can influence results in a positive way for men, or negative for women.

Yes, I've written about this extensively. If you're interested, you can see some of my previous scribblings about research on women and political knowledge here, here, and here. It's an area ripe for more study, if you're fishing around for a conceptual area to work in.  What else skews tests toward men? The topics of questions asked, politics as "game" or "contest" versus problem solving, and using only male leaders in questions versus female leaders (this one makes a huge difference, research says).

In other words, it's a lot more nuanced than presented here.

Okay, this is an interesting study as it includes 10 nations: Australia, Canada, Colombia, Greece, Italy, Japan, Korea, Norway, the UK, and the US. Despite this diversity, "women answered fewer questions correctly than men in every country."  Setting aside the methodological concerns I raised above, there's a powerful media effect in play.  As they write:
"It seems that gaps in exposure to media are related to the gaps of knowledge between men and women," says Professor Kaori Hayashi, co-researcher on the report. He found that the gender-bias of hard news content in all countries plays an important role in gender gaps and underlines the serious lack of visibility of women in TV and newspaper coverage.
So as a media guy, I'm happy to see a media angle pop out.  But I'd love to know the questions they asked, yet I'm having difficulty locating them.  This page shows me only the grant info, not all that helpful, but it's the "report" link from the Guardian piece.

As faithful readers of the literature knows, so much matters in the kinds of political knowledge questions you ask, how you ask them, the kinds of response alternatives are available (multiple choice versus free recall, for example). I'm guessing this study has not gone through peer review yet, so we're not seeing the details. Until then, remain skeptical.  It's hard to see if they used statistical controls and other factors before presenting the gender differences as real.

No comments: