The latest AP poll seems like it has great news for Barack Obama.  After all, it puts his approval rating at 53%, even while scorn for Congress increases.  But how exactly did the AP poll come up with an approval rating ten points above today’s Rasmussen result?  The answer lies in both the survey type and the sample:

The latest Associated Press-GfK poll found that fewer people approve of Congress than at any point in Obama’s presidency. Support has dropped significantly since January to a dismal 22 percent as the health care debate has roiled Capitol Hill. Neither Republicans nor Democrats are safe; half of all people say they want to fire their congressman.

Conversely, Obama’s job-performance standing is holding fairly steady at 53 percent. And over the past two months, the Democrat has gained ground on national security issues, specifically the subsiding Iraq war and the escalating Afghanistan war, as he has spent most of his time — at least publicly — on domestic matters like the economy and health care. On those issues, he still has the support of about half the people.

“I agree with what Obama is trying to do, but nobody is listening to him,” said Grace Pope of Waterville, Maine. But this 75-year-old Democrat added, “I don’t think that the Congress is doing anything.”

Such sentiments and the survey’s results make clear that Obama remains far more popular than House and Senate members as he leads a Democratic Party facing a volatile election-year environment that, so far, seems to be trending in Republicans’ favor. Judging by his standing at this point, Obama seems to be an asset for his rank and file.

An asset for his rank and file?  Hardly.  The AP only reaches that conclusion by surveying adults rather than registered or likely voters, which are polls more suited to predicting outcomes for elections.  The AP knows this, but apparently couldn’t be bothered to adjust its polling technique to more accurately predict electoral behavior — or wanted to put its finger on the scale without too many people discovering its game.

The partisan gap in this sample is another big reason.  While poll after poll shows that gap nearly disappearing, the AP sample has a whopping 11-point advantage for Democrats, 45/34, including independent leaners (page 31).  Without the leaners, it’s 33/23.  Bear in mind that Obama got elected in November 2008 with a seven-point advantage in the popular vote — and that was with significant Republican crossover voting.   The AP’s partisan split at that time was 48/34, 40/24 without leaners.

Rasmussen’s last survey on party affiliation put the gap at 3 points.  Deduct 8 points from Obama’s approval rating, and you get somewhere close to Rasmussen’s approval rating for today.  If the AP actually believes the analysis above based on their polling, I pity them their wake-up call in November.