Are people giving funky answers to questions on Obama's approval to pollsters they perceive as associated with one side of the political divide?
I've been following the TPM averages for the last few days, suspicious of the Rasmussen numbers in light of what seems like a shifting climate on Gallup. I did some tinkering around, and came up with some surprising results. Turns out Obama's numbers improve when you eliminate both ends of the continuum...
Here's the unfiltered TPM numbers:
Now eliminate Rasmussen...
Now, just to be fair, eliminate PPP as well...
That's a difference of three percentage points in just a few days...
Certainly, others are better prepared to assess this than I am -- I've just been surprised to see it go unremarked of late. Obviously, it's dangerous to cherry pick numbers, but can anyone explain this? A difference of three percentage points by eliminating two pollsters identified (fairly or unfairly) with two DIFFERENT parties must be unusual...