I went to Zürich this weekend, and I promise I'll post some pictures soon. It's a lovely city, but unfortunately I caught I cold and currently feel like my head might explode. Anyway, I'm not feeling funny, so I'm going back to politics.
Right now it seems like we might have an Obama landslide. States that I never thought would be go Democrat are being discussed as toss-ups. That's right - I'm talking to you North Carolina. I occasionally hear Georgia mentioned, but I think it's unlikely. I'm just happy that it's a light red in some maps.
So that's all good. But there's talk about something called "The Bradly Effect." This is basically the idea that people lie about who they plan to vote for in order to give a more socially acceptable answer. Here's a link to the NY Times article about it. The name comes from the African-American mayor of L.A. who was predicted to win by a wide margin, but actually lost by a narrow margin.
I'm not really sure what to say about it, except that I hope that it's not gonna be a factor. And really, I think that it's probably not. But I can imagine lying to a pollster to produce a more PC answer. Because this one time I got asked to record my radio listening and send it in for some national polling company. Those of you who know me know that I pretty much listen to NPR all the time. But at the time I was also carpooling to work with people who listened to 96 Rock (and it was while "The Regular Guys" was on - few things in this world are more horrible). I was tempted to lie - because I didn't want the ratings of 96 rock to go up because of me. I also considered lying and including radio that I don't usually listen to, but probably should - like "Democracy Now."
I tell this story because I like to talk about myself, but also to illustrate my point about how a person might lie to a pollster. Because it's not really gonna make that much of a difference, right? And if you were asked by a person who you thought might want some particular answer it would be really easy to tell them what you think they'd like to hear.
So should you get called by a pollster, you should probably tell them who you're really going to vote for. Also, don't say that you're going to vote if you're not. And, in case you were wondering, I ended up being honest about my Radio listening habits.
Subscribe to:
Post Comments (Atom)
3 comments:
Some thoughts on polls.
1) They are lies. The statistical methodologies popular polls use do not use -ahem- robust error measurements
2) I know this because a prof of mine said so. Here's a website he made (for fun, not the views of his employer, blah blah blah)
http://userwww.service.emory.edu/~dlinzer/pres08/
3) Check this out
4) and this was an interesting thing Peggy Noonan wrote:
http://online.wsj.com/article/declarations.html
"People speak of the Bradley effect—more people tell pollsters they will vote for a black candidate than vote for the black candidate. But I have been wondering about the possibility of what may someday be called the Obama effect: You know your neighbors think he's sketchy—unknown, a mystery, "Hussein"—so you don't say you're voting for him, but you are."
check this out (#3) is supposed to be
http://www.ajc.com/metro/content/shared-blogs/ajc/politicalinsider/entries/2008/09/11/alan_abramowitz_bar_bet_obama.html
Also, we may have unprecedented voter turn-out. This is what Team Thomas is counting on to counteract the Bradley effect (i.e., when Kevin told me about the Bradley effect, I said, 'but what about unprecedented voter turnout?' and he said, 'that's a good point.')
Post a Comment