Why Question Order Matters

In a case of studies I wish I had done, the latest Journalism and Mass Communication Quarterly includes a research paper that finds that question order can influence how interested survey respondents say they are -- the good old political interest variable.

What decreases political interest?  Asking them political knowledge questions beforehand.

That seems kinda obvious, now that I see it on paper in a major journal, but I'd never given it much thought until now.  The study, by Dominic Lasorsa, is a nice and tight study that examines not only how asking political knowledge questions might affect political interest, but also neatly examines the placement of questions that might excuse one for not being very knowledgeable.  It all boils down to priming and context effects. 

The placement of political knowledge items in a survey before a political interest question had its greatest impact among women, the less educated, those of lower income, and younger respondents.  These are folks who typically, for various reasons, do less well on political knowledge tests.  So being hit by questions that make you feel less knowledgeable leads to you saying you're less interested in politics.  There's some priming going on, but also probably some self-esteem.  Blowing some knowledge questions makes you a little more likely, I'd think, to then say politics doesn't really matter to you.  Ace 'em, and sure, politics matters.

This latest issue of JMCQ (Autumn 2009, Volume 86, Number 3) includes "Political Knowledge" on the cover six studies inside that touch on some way political knowledge.  I'll try to discuss them as the week progresses, assuming I actually get prepared for my Spring Semester classes.