Being somewhat of a statistics geek I enjoy surfing the sites for poll data. With every new election year there seems to be more and more polls done more and more often. The Zogby poll has shown tight polling numbers and the Gallup poll has shown a consistently higher lead for Obama. The “poll of polls” being done by Real Clear Politics makes an average by combining several different polls. The pollsters all seem to have their formulas for coming up with their numbers. Some are weighted by “likely voters” which is often based on whether the person being polled voted in the 2004 election. Some pollsters squeeze the younger vote polling data to adjust for historically low turnout on Election Day. I place more confidence on the pre-election polling data than on exit polling. In 2000 I was on my way to the polling place when I heard the news sources call the race for Al Gore based on a combination of the early returns and the exit polling coming out of Florida. Only a small percentage of the Florida votes had been tallied at the time. By the time I got home and was cracking my celebratory beer, the election results had already begun to unravel. As a higher number of election results came in the newscasters became increasingly nervous about their earlier projection. At some point in the evening, the analysts removed the blue hue from Florida and took the Electoral College numbers away from Gore. My celebratory mood subsided as I changed my beer for something more expedient. Despite my lost confidence in polling data I am willing to make a prediction about the outcome in tomorrow’s election. Obama trounces McCain by at least 7 percentage points and crests the 270 Electoral College threshold by at least 60. I have to admit, this year is feeling a bit like Christmas. I just hope that I don’t get another stocking filled with coal….
Rather overdue, here's the link that shares the fears of HAVA that Robert Kennedy and I had. I agree a lot with Brian thinking about about the validity of pre-election polling (as I say this after the fact, hah) because I would think that there is a strong intention to vote in this pivotal election, and there should be honest responses provided for surveys that provide such important statistics. But "likely voters" aren't going to be the only demographic pulled for all of these surveys...it's too bad that ALL registered voters aren't considered likely voters.
Posts
2 comments:
Being somewhat of a statistics geek I enjoy surfing the sites for poll data. With every new election year there seems to be more and more polls done more and more often. The Zogby poll has shown tight polling numbers and the Gallup poll has shown a consistently higher lead for Obama. The “poll of polls” being done by Real Clear Politics makes an average by combining several different polls. The pollsters all seem to have their formulas for coming up with their numbers. Some are weighted by “likely voters” which is often based on whether the person being polled voted in the 2004 election. Some pollsters squeeze the younger vote polling data to adjust for historically low turnout on Election Day.
I place more confidence on the pre-election polling data than on exit polling. In 2000 I was on my way to the polling place when I heard the news sources call the race for Al Gore based on a combination of the early returns and the exit polling coming out of Florida. Only a small percentage of the Florida votes had been tallied at the time. By the time I got home and was cracking my celebratory beer, the election results had already begun to unravel. As a higher number of election results came in the newscasters became increasingly nervous about their earlier projection. At some point in the evening, the analysts removed the blue hue from Florida and took the Electoral College numbers away from Gore. My celebratory mood subsided as I changed my beer for something more expedient.
Despite my lost confidence in polling data I am willing to make a prediction about the outcome in tomorrow’s election. Obama trounces McCain by at least 7 percentage points and crests the 270 Electoral College threshold by at least 60.
I have to admit, this year is feeling a bit like Christmas. I just hope that I don’t get another stocking filled with coal….
Brian
http://www.msnbc.msn.com/id/3036789/%27)
Rather overdue, here's the link that shares the fears of HAVA that Robert Kennedy and I had. I agree a lot with Brian thinking about about the validity of pre-election polling (as I say this after the fact, hah) because I would think that there is a strong intention to vote in this pivotal election, and there should be honest responses provided for surveys that provide such important statistics. But "likely voters" aren't going to be the only demographic pulled for all of these surveys...it's too bad that ALL registered voters aren't considered likely voters.
Post a Comment