September 04, 2004

How 'Bout Those Polls?

sep2ndpolls.gif

I just wanted to say something about today's strange Time poll. If you look at the three last polls taken, it is outrageous to believe that the first two got it that wrong. It is also hard to believe that the Time poll could be a that much of an anomoly. I am prepared to concede that Bush got a bounce. I cound even accept a five, six, or even a seven point bounce--but eleven points is far too much of an anomoly to believe.
It is also alarming that Time magazine makes no attempt in either their statement or methodology or in their preface to the poll on their website about the extreme differences between its poll and the other recent polls. I think it should be the responsibility of the polling group to state that, even though they can not dispute the numbers they received, they should concede that it is statistically inconsistent with other polls taken during the same time period.
And if you think this is just sour grapes on my part, it is not. I would be skeptical if Kerry got a double-digit lead in a poll if several others polls showed him with marginal or no gains. However, I would also be excited, and then when the other polls started to show that the double-digit poll was an anomoly, well, I would be pissed off. And trust me there are a lot of happy Republicans this weekend, but the party will crash by the beginning of next week. Because even if Time magazine won't do their job, then I will.

Time Magazine's Poll is a statistical anomoly. And it should not be taken seriously.

Posted by Paul Hina at September 4, 2004 12:22 AM