Pollster.com

Could you guys talk at some point about polls such as this where the demographics seem off?

1. Are 50% of self-described republicans really women?

2. Are 18-34 year olds really only going to make up 12% of the electorate?

3. And the biggie....Should it set off any alarm bells when the sample disparity between respondents who claimed they voted for Bush in 2004 versus Kerry is 13%? Their numbers are 48% Bush, 35% Kerry, and 13% don't know/refused. So are 92% of those 13% who refused to answer all Kerry supporters (which after partitioning gives a result of 49% Bush / 47% Kerry)?

I just wonder if some of these pollsters haven't picked up on the possibility that a lot of currently self-described independents used to be republicans but switched out of disgust with Bush. And many formerly self-described independents are now self-described Democrats. Wouldn't this skew some results towards a more conservative sample?

Very good news for McCain. Obama's post-Hillary concession bump is small according to this poll.

This poll is actually good for Obama because the total sample is people who voted Bush on 04 by 13% even though in 04 Bush von by under 3%. So the sample is skewed heavily to the right and Obama still wins or most importantly only 35% of the sample voted Kerry yet 44% go for Obama and 48% of sample voted Bush and only 42% go for McCain.
Per 2004 results (if samle was only Bush Kerry voters and properly weighted) then Obama would lead 55.2% to Mccain's 44.6%

@ kingsbridge -- two delusional posts in the same day?

"Obama's post-Hillary concession bump is small according to this poll."

Fine, except you forgot to mention that the dates for this poll were June 5- 8 and Hillary's "concession speech" was the 7th!

Again, just saying something doesn't make it true. Just because the poll was released yesterday doesn't mean it was conducted this week. As we see, it was conducted mostly BEFORE the concession speech. Ugh, you can't make this stuff up.