If the NEP folks stopped polling an hour before the polls closed, then what proof do we have, if any, that the early exit polls missed late-voting Republicans?
Posted by: Steve Soto | Nov 29, 2004 5:04:48 PM An article came out tonight that the Texas Hispanic vote estimate has been revised from 59/41 Bush to 50/49 Kerry, which could explain the drop from 44 to 42 nationally for Bush. http://story.news.yahoo.com/news?tmpl=story&cid=694&e=2&u=/ap/20041129/ap_on_el_pr/eln_texas_glance_corrective Posted by: Matt | Nov 29, 2004 9:09:49 PM Mark, I’m very intrigued by how the final weighting is done to bring the exit poll results in line with the actual numbers. Is it nothing more than weighting by turnout in the sampled precincts compared to the “rest-of-the-world”, and then hoping that the results then line up? If that’s the case, then I’ll be much less skeptical about exit polls. But I suspect that the weighting schemes are more complex, and that the actual results must be factored into the weighting, too. If that’s the case I think the entire validity of the exit poll data is somewhat suspect. Here’s a toy problem example to illustrate the poten
Related Questions
- If the NEP folks stopped polling an hour before the polls closed, then what proof do we have, if any, that the early exit polls missed late-voting Republicans?
- Who makes the decision that school should be closed, open late, or dismiss early in bad weather?
- Should political polling be banned during an election year? Why or why not?
