Friday, September 26, 2008

Men, Women, Cellphones and Sample Bias

I was reading a an analysis of why Ann Selzer is such a good pollster over at 538 and got to thinking about sample bias in political polling. Young people tend to be more cell phone dependent as everyone has read by now. This makes them less likely to to be counted resulting in the polls skewing older (and more favorable to McCain). Young people are more likely to screen their calls and not pick up unrecognized numbers. Everything so far is old news.

My thought is this, aren't men much more likely to respond to an unrecognized number. My personal experience suggests that men often pickup unrecognized numbers and women rarely do. Do other people agree that this is the general trend? If so, then there should be a bias towards including more young men than women. The cross-tabs would clarify this. If it's a problem then, since women are more likely to support Democrats, this would hurt Obama's poll numbers relative to his actual performance.


At 9/27/2008 , Blogger Kol Ra'ash Gadol said...

not necessarily - it depends a lot on what you do for aliving. I answer unknown numbers all the time, because I could never program in every number i need to know...


Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home