That Bradley effect

It's when people tell pollsters what they think they SHOULD say but then vote their prejudices. Look at this from a Slate article:

1) Polipundit suggests Obama may have benefitted in Iowa from a "reverse Bradley effect.' The open, public voting of the caucuses provided Democrats with

"a golden opportunity to show your next-door neighbors just how enlightened and progressive you are, by supporting the liberal black candidate."

On a secret ballot, Obama wouldn't do as well. If the Reverse Bradley Effect holds, then, Obama will do worse in New Hampshire than his Iowa triumph would lead you to expect, even if Hillary does nothing to change anyone's mind. ...

2) I haven't heard any MSM pundit mention another possibility a Polipundit reader mentions: that Romney may have done worse than the polls indicated because the Republican caucuses did use a secret ballot--and people who wouldn't tell a pollster they weren't going to vote against a Mormon in fact voted against a Mormon. This is not a reverse Bradley Effect. It's the regular ol' straight Bradley Effect;

Certainly makes sense to me.