Time's statistician weighs in on the NYT poll article

I noted earlier today in Reality Bites that NYT's sloppy journalism had struck again.

TPM Election Central has the story on NYT's latest goof-up. Seems that Adam Nagourney had a tale to tell and didn't let the facts in the poll (pdf) deter him. The Obama campaign looked at the reported data and identified a more than a few problems with Adam's narrative.

Now, via Karen Tumulty, Time's statistician Jackson Dykman has weighed in with some observations on the NYT article versus the poll.

I too was really struck by the NYT's characterization of its own poll. I read the full poll first and then the story when they posted it last night. I've rarely seen a story so wildly off from the actual data on which it is based.

Aside from point C in the Obama response (which is true and basically negates the story), the premise of the story is, well, utter nonsense. [...]

I can't decide whether it's the headline or the story that really creates the schism. The hed in the paper is: "Poll finds Obama candidacy isn't closing country's divisions on race."

Are you kidding me? The guy just wrapped up the nomination. Racial divisions in the U.S. have a wee bit of a 400-year head start on him. If Obama goes on to win the election, I really hope the Times does this poll again in four years. Whatever the result, THAT would be a story. [...]

Two other points. In the poll data, 79% of white voters think an Obama administration would treat both whites and blacks the same. And 82% of white voters think a McCain administration would treat both races the same. Okay, fine. The real story in this question is this--90% of black voters think Obama would treat whites and blacks the same, but only 50% of black voters think McCain would treat both races the same.

Yet Obama is the one who's failing to close "the country's divisions on race"?

Finally, I looked it up again because I couldn't believe it, but Point E in the Obama response is correct. Why on earth would the story say "there's even racial dissension over Mr. Obama's wife, Michelle: She was viewed favorably by 58% of black voters, compared with 24% of white voters."

The numbers for Cindy McCain: 20% favorable among white voters, 9% favorable among black voters (!!!)

I've long made the argument that journalists care wildly more about candidates' spouses than readers do, but someone needs to tell me why the racial dissension is "over Michelle Obama."

Thanks to Ms. Tumulty and Mr. Dykman for also weighing in.

Adam Nagourney does have a track record of deciding what the CW is and then writing his article to fit. He provided no end of frustration in his coverage of the Lieberman vs. Lamont race. This is just more of the same.