• woody

    For decades, exit polls (when done on a large scale, at multiple time frames, at multiple precincts) have been very accurate. The reason is that you’re not asking who someone MAY vote for, you’re asking them what they just did, moments ago.

    There are factors that can throw them off, but throughout the world exit polls have matched very closely to the actual vote when conducted properly. The rare instance where this doesn’t work tends to be when there are extreme factors involved. Early voting, high amounts of mail-in voting, and the “Bradley Effect” are good examples that may affect exit polls in this race.

    Reality is tough that for may years, exit polls worked well. I think the past few cycles they’ve failed in part due to the polarization, the lack of resources (or slant) of groups conducting them, and that there was legitimate voter fraud going on. (And by this, I mean voters were being defrauded by having their votes shifted by machines or officials moving ballot cards into the wrong stacks. Not people trying to vote more than once.)

  • The Bradley Effect appears to have not existed. Pollsters went back and looked at the poll Bradley did the night before, and found it was flawed.

    Studies done in 2006 in elections with Black and White candidates, including races in the South found no Bradley Effect.

    Yes, it exiting polling is done on a genuinely random sample, then it’ll be accurate. But it appears it’s hard to do that.

Powered by WordPress. Designed by WooThemes