Nate Silver: The polls aren’t wrong
Bias? The stats guru who nailed the 2008 election tells Salon it’s very hard for this many surveys to all be off
http://www.salon.com/2012/09/29/nate_silver_the_polls_arent_wrong/
The most interesting revelation to me in this article is that in the battleground states Romney has never led and indeed has trailed in nearly all the polls ever taken. Rather convenient that the media ignored that tiny thing and trumpeted the statistical heat at the national level (one has to wonder if Romney won the popular vote, but lost the electoral vote, as Gore did, would the GOP now complain about the electoral system?), but I have read reports that make plausible arguments that the media deliberately keeps the story on how close the race is in order to promote viewership.
Of course, as the subject of the interview points out, favoring Obama by 80% means that one out of five times the opponent (Romney, in this case) is _expected_ to win.
Also interesting was the comment that more than a certain number of polls starts to diminish the aggregate value as it appears pollsters don’t want to stand out and will use statistical methods to bring their results into agreement with the mainstream. It appears that due to the statistical nature of polls one should expect the results to _diverge_ rather than converge as is often seen.
Of course, like any other area of inquiry that depends on a deep understanding of math and statistics, I am quite sure that the majority of pollsters are hacks just trying to squeeze a few bucks out of credulous partisans. It appears clear to me that any deep understanding of polls immediately takes one out of any partisan world, something clearly incompatible with politics.
There’s an old saying that there are lies, damned lies, and statistics. It’s an improper view of the fact that there are liars, damned liars, and statisticians who either don’t know their field, or manipulate it.
Having delved into statistics as a means to ensure quality, reduce costs, and increase productivity, I was amazed at the methods recommended by “experts” to “fix” unfavorable outcomes.
It’s asinine to ask a question again simply because you didn’t like the first answer.