In late September, the buzz was a theory that polls were massively skewed towards Obama because they failed to poll equal numbers of Democrats and Republicans. One web site, UnSkewedPolls.com continues to publish results showing Romney way ahead using its own method of adjustment to compensate for the fact that most folks consider themselves Democrats.
In a major article, “Obama stays ahead – just”, at YouGov.com, Peter Kellner talks about the issue of adjusting polling samples for demographics and party affiliation, and he discusses in some detail why the YouGov polls differ from most national polls, and in particular why YouGov didn’t show a huge swing towards Romney after the first Presidential Debate, a much different result from other polls.
YouGov’s methodology differs from other pollsters in that they have a large panel of individuals that they have tracked over time. A typical telephone poll, for example, asks if you are registered to vote and takes your word for it. YouGov has checked public records and knows for sure whether 150,000 members of its panel are registered to vote.
It’s a substantial article and I won’t try to summarize it here. If you’re a polling junky, check it out.
Today the YouGov Vote Intention poll stands at 48% Obama and 46% Romney.
Polling is important in the face of the ramp-up of conspiracy theories, for example at WorldNetDaily, predicting massive voter fraud and results tampering in the upcoming election.