The conservative blog HotAir has an interesting methodological criticism of the ABC News/Washington Post poll released earlier today:

Let’s deal with the sample first. The poll includes 35% Democrats, 25% Republicans, and 34% independents. That would make sense — if Obama had won the election by 14 points. He won it by seven, with independents and some Republicans breaking for Obama. Here’s a hint to pollsters: if you’re tracking a bigger gap between Democrats and Republicans than we saw in the presidential election, your sample is almost certainly unrepresentative of the electorate. At least it’s improved for WaPo/ABC, which had it at a ridiculous 13-point gap in June.

The skew explains why Obama still has a 57% approval rating in this poll when Rasmussen and Zogby have him under water and Gallup has him closing in on 50%. Even with the skew, his disapproval numbers have risen sharply, especially on ObamaCare, which is practically all anyone discusses. Half of the survey respondents oppose it, with 40% strongly opposed, compared to 45% approving and only 27% strongly approving. Obama is losing the argument, and the longer it continues, the more his overall numbers will erode with it.

WaPo/ABC sees independents as the key, as should Congressional Democrats facing elections next year. At the beginning of 2009, Obama had high ratings among independents, but no longer. Only 41% now trust Obama’s judgment, down from 61% in January. They disapprove of Obama’s handling of health care by 57%, almost all of whom strongly disapprove, and the same number oppose the policy itself. Among the half of independents that say ObamaCare will affect their vote in the midterms, they swing 2-1 to saying that they will oppose candidates who backed it.