Comments for Are Surveys Scientific? at http://www.randi.org/site , comment 1 to 2 out of 2 comments http://www.randi.org/site Sat, 25 May 2013 09:55:00 +0100 FeedCreator 1.7.3 http://www.randi.org/site/index.php/swift-blog/1862-are-surveys-scientific.html#comment-26021 ...Miss the point. The idea behind election survey is to predict the election results by noting who will be in the polls. One cannot simply sample people randomly or else one will sample minors, non-citizens, etc.. So one MUST over-sample some groups and under-sample others. The question is which groups are over-sampled in what way. It seems that most polls over-sample based on the latest available relevant data: the 2008 federal elections. Yes, there were more recent elections -- i.e., for Congress in 2010 -- but voting patterns for congress differ from those of the federal elections. Can these polls be skewed and nto represent reality? OF CORSE they can, if there is a change from 2008 to 2012. Polls can be wrong. But serious pollsters cannot invent the polls they *guess* represent the *unknown* will of the electorate right now. They need to poll by factual data they have. - Skeptic Fri, 05 Oct 2012 21:13:22 +0100 http://www.randi.org/site/index.php/swift-blog/1862-are-surveys-scientific.html#comment-26013 Years ago I was taking a preference survey regarding auto design. After clicking boxes at a computer for an hour and a half, I got tired. I know it's wrong but I started liking every design. Ergo, the Pontiac Aztek. My fault. Sorry. - garman Thu, 04 Oct 2012 07:36:36 +0100