Tuesday, September 25, 2012

opinion poll truthersim

like dave weigel, i have also become fascinated with opinion poll trutherism. dean chambers decided that the polls he was seeing "just didn't look right" and so he reweighed the polls giving republicans a 3% advantage in the sample. voila, suddenly mitt romney is comfortably ahead in every poll!

the polls could be wrong. certainly individual ones are. that is why i like to rely on poll averages, which should smooth away the outliers and weaker polls. while i think it is unlikely, it is also possible that all polls, or enough polls to affect the average, are wrong too. but there's no way to tell whether they are wrong right now. the only measure of opinion prior to an election is polling.. the real test will come when we see how the polls stack up on election day. until that happens and we have actual data on the polls' reliability, the feeling that something doesn't look right is just a reflection of your own hopes and biases.

i wonder what will happen if the standard polls turn out to be right. will mr. chambers admit that his "unskewed polls" were, in fact, skewed? i'm guessing not. either he will argue that the artificially pro-obama polls of the MSM affected the voter results, or he claim the election was itself skewed by vote fraud. and thus the rightwing bubble forever closes in on itself. and the right will drift even further afield from reality as experienced by most of the country.