New York Times skews poll results

Timeswatch reports the following about a recent NYT poll:

But the Times doesn't mention in its story that the public perception of Bush's handling of Katrina has actually improved this week, from a 20-point gap in a CBS poll a week ago (38%-58% approval-disapproval) to a 6-point gap in this latest poll (44%-50%). For that tidbit you have to dig into the poll questions online. (It's Question 8.)

Only in the fine print of a sidebar paragraph does the Times admit its poll is skewed: "Black Americans were sampled at a higher rate than normal to permit the analysis of black attitudes in greater depth." 211 out of 1088 (almost 20%) respondents were black, compared to the 13% share of the population that is black. The story doesn't bother mentioning this, although black voters skew Democrat, which would affect Bush's polling numbers.

What's next, oversampling New York city voters?