Sunday, January 15, 2006

What The Post-Standard Didn't Tell Their Readers

Today's Syracuse Post-Standard opinion page featured results from an AP-Ipsos poll conducted January 3-5, showing that Republicans stand to lose Congress by 13% (the P-S is coy about their online presence; I can't find a link to the story). The P-S was careful to include everything anyone could want to know about the poll in order to judge its validity:
The Associated Press-Ipsos poll on public attitudes about President Bush and Congress, conducted Jan. 3-5, was based on telephone interviews with 1,001 adults from all states except Alaska and Hawaii. Results were weighted to represent the population by demographic factors such as age, sex, region, race and income. The margin of error is plus or minus 3 percentage points.
Well, except for one small detail. The poll (available here in Adobe Acrobat format) respondents were 52% Democrats and only 40% Republicans.

This is part of the same tainted poll I discussed earlier here. There's no way to tell whether the editors simply trusted AP-Ipsos to get it right and got burned, or went with the poll "results" despite the grossly biased sample.