NBC/Wall Street Journal Embarrassed on Oversampling updated
 |
… .. But one which, I think, tends to undercut the arguments of the Obama campaign that he would be a substantially stronger candidate in the general election than Clinton. And evidence that the Rev. Jeremiah Wright has taken the luster off Obama and made him a weaker candidate against McCain than he was two week ago. … ..
This poll has gotten them into trouble, so they’re trying to explain themselves. Short answer: You all are idiots. In a post entitled “What ‘Oversampling’
Means,” the group that conducted the poll “explains.” In other
words, covering their political posterior for a poll that should never have been
released they blame you for being ignorant.
As you know, the sample for the March 24-25 poll on race included an “oversample”
of 100 African American voters. – MSNBC
First Read
Here’s an interesting comment from reader Dr. TS:
I literally just received an E-mail from my husband on the innacuracy of
this poll. He is more statistically inclined. He said the following:I’m surprised they released this poll. The actual results mention
the over sampling but it doesn’t match up with the question about
race. This shows 11% were AA which would be 77 out of 700. But, they state
they oversampled by 100 which means 177 were AA (25%). I don’t see
how they can make any inferences about overall voters since the sample sizes
aren’t reflective of the actual population. They don’t say what
they did with the extra 100 AA respondents. Plus, the AA respondent results
have a higher error (since it’s a smaller sample size) so you really
can’t say anything about those results. (This poll’s results has an
error of +/- 7.4%!]
I’ll let you math and poll experts break it down further, but it’s clear that
the poll MSNBC is pushing, compliments of Clinton haters Matthews and Olbermann,
isn’t worth the paper it’s printed on. They’ve even got a representative of the polling company on MSNBC touting how good the poll is for Obama, and how bad it is for Clinton.
Bottom line: Oversampling Asking African Americans to find out what they think
on Obama’s pastor disaster isn’t really the issue is it? With African Americans
backing Obama 9 to 1, does anybody really believe that Wright is going to move
them? What about how white voters feel? That’s the problem with the poll, especially
when you look at other evidence that directly refutes the bogus NBC/Wall Street
Journal nonsense.
Clinton’s numbers have become somewhat more unfavorable. But the major difference is that McCain’s fav/unfavs have become marginally more favorable, while Obama’s have become significantly less favorable.
But again, they think everyone will just fall in line and
believe this baloney. Instead, NBC/Wall Street Journal have been busted.
However, all this is really now a moot point, because Rev. Wright has offered more
nuggets of his nuttiness, which I’ll cover in the next post, with cable running with Obama’s pastor disaster again today. Obama’s not out of
this one by a long shot. That’s why his camp, with help from Nancy Pelosi, is
trying to shut down the primary fight.
Two words of advice for the DC elites:
Back. Off. Otherwise they’ll create a bigger problem than they have now with
Clinton voters, if that’s possible.
UPDATE: I firmly believe that the NBC/WSJ poll is incorrect and inaccurate on its broader assumptions. Pollster.com believes the poll to be accurate. In addition, I received this from a reader, which I gladly post as another voice that the NBC/WSJ post is accurate:
I enjoy your blog very much, but (putting my PhD from MIT to use
here) I have to respectfully suggest that your post about the NBC-WSJ poll warrants a correction. Sampling means the number of interviews done for a group – not the percentage of responses included in the overall result. More than 11% of the survery-respondents queried were African-Americans, but the overall result was weighted to represent only 11% of that over-queried African-American group. This really is a very acceptable way to increase the statistical reliability of a poll, particularly with groups that represent only a relatively small percentage of the overall population. – Respectfully, MdeV
Comments are closed.