Allow me to get irritated for a minute. The first article below (first link at the bottom) is an example of inciting unnecessary fear.
The so-called “Bradley” (a.k.a., Wilder, a.k.a., Dinkins) effect, of which there is sparse evidence, has been assigned to Obama for one reason: because he’s an African American candidate (and race and drama go hand in hand). Polls by their very nature are “retrospective” (snapshots of yesterday) not “predictive” (forecasts of tomorrow).
Most in the media who are talking and writing about a possible Bradley effect with Obama have very little theoretical understanding of survey research methods, polling, or research on racial attitudes in America. Moreover, there are plenty of times when black candidates run and there is no pre-election poll versus outcome discrepancy (see 2006, with Swann, Steele, Ford, and Patrick).
In fact, most of the differences in poll results before an election and the outcomes of an election are due to factors like bad polls (conducted improperly), margin of error + the undecided voters not being considered correctly, the different polling methodologies (i.e., phone vs. web vs. in-person vs. mail vs. interactive voice response), and other survey context effects (i.e., question wording, question order, interviewer effects, days/times of the week, etc.).
I won’t even get into all the other errors that are known to influence response: coverage error (is the call list representative?), non-response error (are the people who didn’t answer different than those who did?), and measurement error (are we really measuring the behavior we’re talking about?).
Trust me when I say, if there is any effect to “look out for” in this election it is the underestimate of young people turning out, and the lack of coverage for young people (who’s numbers are often weighted up) in polls.
This is not to say that people are completely truthful when they answer polls, we know they are not. But, this “ain’t” a single state (CA) election in 1982, or 1988 (Wilder), or 1990 something (Dinkins), this is a national sample of people, some of whom have plenty of reasons to say they wouldn’t vote for Obama, other than race. Please, don’t believe the hype.
At some point in our lives we’ve all been told the Boogey man is under our bed, and every time we look…..we see nothing. This is because 1) the boogey man isn’t real, and 2) if he was, why would he hide under the bed? The point is, when you look for the wrong thing in the wrong place, you wind up believing “it” could still be there.
Below are some articles (for bedtime reading only) to think about.
There is a Bradley Effect
http://www.cnn.com/2008/POLITICS/10/13/obama.bradley.effect/index.html
There is a “reverse” Bradley Effect
http://www.politico.com/news/stories/0508/10397.html
Moving beyond the Bradley Effect
http://www.cbsnews.com/stories/2008/10/13/opinion/pollpositions/main4519166.shtml
Age and Cell Phones-Effects on Obama vote
http://pewresearch.org/pubs/964/
– David Wilson