Real Clear Politics http://www.realclearpolitics.com/ reports the average of major polls in Texas and Ohio were breaking towards Hillary Clinton on the eve of the Democratic Primary Contests. Click and check out the charts and graphs under, "Clinton Squeaks Ahead in Texas," and "Late Ohio Polls Break for Hillary."
Will these polls be accurate?
The AP reports that record early voting in Texas -- which might make up anywhere from 1/3 to 1/2 of Texas voters -- might have the election half-over in the Lone Star State. Early reporting might be a good indication in how Texas and the polls will go tonight.
http://ap.google.com/article/ALeqM5i9EB1bZOa3bSVHoLh75KNClKet6AD8V68GR00
Like the undercount of cell phones in the 2004 presidential election, early voting has been problematic for polling services this campaign.
The most notable of the inaccurate polls predicted Barack Obama as the big winner of the New Hampshire and California primaries.
But the cause of the polling errors might have more to do with the big changes we have seen in these elections rather than big changes in public opinion.
To accurately predict whom voters will support on Election Day, a pollster must first determine who will be a voter. Surveys of “likely voters” typically define likely in many different ways.
Besides asking if a respondent will vote on Election Day, they might also ask a series of filters: Have you voted in the last two elections? Have you ever voted in a primary? Have you ever not voted? So when voter turnout is very high, as it has been this primary season, the poll might prove to be inaccurate because so many unlikely voters actually turn out.
The biggest change that has affected the accuracy of opinion polls seems to be the new phenomenon of early voting.Nearly 1 million Californians used early voting before the Super Tuesday election.
Polls that took a snapshot of popular opinion in the days before the election might not have been able to gauge those opinions that had been cast as votes weeks ago. This is yet another reminder that, while polls are crucial in feeling the pulse of American opinion, the Election Day ballot box still remains the most important and most accurate electoral survey.
8 comments:
Polls are completely unreliable in this election, and are becoming so more and more in modern elections. Polls often say one thing, yet the other can happen - the reality is, that polls are only samples, and especially because young people aren't included (because of lack of land lines and such), they will predict nothing. All they do is show general trends. All's fair in love, war and politics. Anything can happen, and so is the beauty of the American electorate. I think that Obama wins Texas by a small majority, and loses in Ohio. He cuts it close in Pennslyvania, and Clinton lives to see another day.
People shouldn't put too much stock in polls - it's nice to have an idea of public opinion and different trends, but it's very hard to conduct a good, unbiased poll. Despite extensive polling, there have been a lot of surprises in the campaigns so far, and there are undoubtedly more to come.
I am not quick to throw the polls out with the bath water. I do believe most of the polls have been accurate this primary season, that's why I like Real Politics, which takes the average of all the legitimate polls.
But the "likely voter" designation and the phenomenon of the massive early vote are trends the pollsters may need to re-examine.
I wouldn't advocate throwing out polls, but I would suggest taking them with a grain of salt. I should also probably clarify my statement about surprises: I meant that public opinion can shift quickly, invalidating previous polls.
Vivi is right, especially when one considers New Hampshire where Hillary picked up many late undecided women voters. Because those undecideds went decidedly to Hillary the weekend before the primary, it made it look like the polls were flawed. They really were not wrong.
Maybe it was just the pundits who couldn't comprehend a high percentage of undecideds going the same way.
It will be interesting to try to figure out how undecideds break tonight.
Could you really be undecided at this point?
While I was doing homework in my room last night, I heard my dad closely paying attention to the voting in Texas and Ohio. The trend seemed that votes for Clinton were quickly catching up to votes for Obama in Texas, and that votes for Clinton were highly dominating in Ohio. Did Clinton's comment on Obama's shifting position on NAFTA issue make difference in the voting result for Ohio and Texas, and was her comment accurate?
I completely agree with Vivi. The polls are an excellent resource, however it is completely unreasonable to expect them to be very accurate. For example, in Texas, a lot of people claim that Hillary's huge late, undecided voter margin was due to the phone ad. The polls are subject to change because public opinion changes so fast.
I pretty much agree with everyone in regard to polls. Sure, they're a pretty good gauge of public opinion, but nobody really knows what they think until they actually cast a ballot. People can change their minds at the last minute. I think we're too obsessed with polls.
Post a Comment