Saturday, November 3, 2012

A Tale of Two Electorates (Part IV)

Here Niall Stanage of The Hill tells us about two different data sets reporting on the ethnic makeup of the 2008 electorate.  The first and most often quoted is a set from 18,000 interviews by media polling organizations at the exit door of the polls presumably on election day.  The second is a set from 60,000 interviews by the Census Bureau, presumably weeks or months later.  About the first, note that election day voters were about 40% of the total in New Mexico in 2008, while regarding the second, voters memories are notoriously faulty.

Here are the two ethnic breakdowns compared:

Thursday, November 1, 2012

A Tale of Two Electorates (Part III)

My son Matt got wound up on the national vs. state poll question reading Dan McLaughlin, who writes the Baseball Crank blog.  If you are a baseball fan, a fan of Michael Lewis's Moneyball, or you've ever heard of the baseball uber-statistical methods known as Sabermetrics, you'll like this almost as much as Matt does.

It was number four with a bullet -- as Variety used to say -- on today's late edition of RCP.

A Two of Two Electorates (Part II)

The picture is becoming clearer.

Here is Jay Cost -- once an election analyst with RCP, now with The Weekly Standard -- pointing out what should have been obvious weeks ago: that the two different kinds of poll results for Romney v. Obama, whether in individual states for incorporation into an electoral vote model, or nationally for estimating the popular vote, are the result of sampling different populations.

[Oops!] A Tale of Two Electorates

Out of lack of familiarity with Rasmussen's site -- I've not been following it -- I picked up Pennsylvania's numbers instead of Ohio's and concluded that Ohio wasn't leaning Romney.

Rasmussen's national and state numbers are much more consistent than the RealClearPolitics national and swing state averages are, just as they should be, having been drawn from the same data sets and analyzed by the same techniques.

More to follow.