These are the "Big Q" results reweighted to reflect actual voter registration data:
This sample is quite "off" but in erratic ways. Some of the erraticness cuts one way, some of it the other way.
A lot of things are wrong with this poll.
There are too many blacks, not enough hispanics and whites, too few women, and the sample is a tad too democratic. Nothing is "hugely" out, but just about everything is out enough for me to get jumpy on trying to read anything into this one.
I guess my comment on this is that it is a Uni-poll, done in the Summer, by trainee student pollsters, in a state the Quinnipiac has not polled before, in a State that is a real total ^&^$%@!! to poll.
When I reweight everything for gender, race, partisan ID etc, I keeps coming out either even or Bush +1, in the head to head (except for male/female which goes the other way) so if I had to I'd call this poll a tie, but I'd rather just toss this poll out all together.
Jeb Bush at 46% approval is waaaaay low which makes me "guess" this sample leans left, but there are too may mismatches going both left and right to be certain of anything.
Quinnipiac has been a tad erratic this year. They had a Pennsylvania sample about 3 months ago that was a real mess also.
Where is Mason-Dixon when you need them ...?
