When in doubt, make stuff up

Hey, kids! How does the Kenyan Muslim commie's approval rating in this month's Fox News poll compare with last month's?
Overall, 42 percent of voters approve of the job Obama is doing as president, while 53 percent disapprove.  That’s mostly unchanged from December when 41 percent approved and 53 percent disapproved.

"Mostly unchanged" is a pretty good description of a 1-point increase in approval, given the sample size (1,010 registered voters). The most likely thing it represents is a slight increase in the population value, though just because that's the best bet doesn't mean it's a very good bet. A well-run poll (give Fox credit for consistency in sampling) is sometimes going to find a slight increase in the sample even when the population value declines. That's why we report measures of sampling error -- and why even Fox reports that that approval number "remain(s) near record lows," rather than cratering upward. Fox is doing basic, up-the-middle framing; The Fox Nation is just making things up, because even if you get far enough to find out you've been lied to, you're feeling so good that you really don't mind. Because BENGHAZI!!!!

Brief detour. Polls make news, broadly, for one of two reasons. Sometimes editors think a poll represents something you don't know about public opinion. Other times, they think the poll tells you why everything you already know is still true. You might have noticed over the past few months that Fox has been assiduous in reporting Gallup's results, which tend to have the president in the quite low 40s, but hasn't made much of Rasmussen, traditional darling of the foamy-mouthed right, which has him in the upper 40s. We could plausibly hypothesize that Fox is deeply interested in public opinion, as long as the public has the right opinion.

Again, Fox has generally been circumspect in reporting on its own polling results, and it's good about making the details public. That makes its data on straightforward topics like presidential approval and party affiliation pretty valuable, but it also up being a nice window into what Fox really wants to find:

Majorities of American voters blame former Secretary of State Hillary Clinton and President Barack Obama for what happened at the U.S. consulate in Benghazi.

Which is what the numbers say. I mean, it isn't Fox's job to ask respondents who they blame for Benghazi, rather than "As head of the State Department at the time, how much do you blame former Secretary of State Hillary Clinton for what happened at the U.S. consulate in Benghazi?"

And, of course, if you ask people whether they think the Kenyan lied to them, you might get something different than asking why they think the Kenyan lied to them:

In the aftermath of the attacks on U.S. diplomatic facilities in Libya, the Obama administration falsely claimed it was a spontaneous assault in response to an offensive online video. ... Which of the following do you think best describes why Obama administration officials gave false information in their early public statements about the September attacks in Libya?

It bears repeating -- say, could that be the top story on the Fox homepage from Sept. 12, 2012? -- that the claims in question were largely the ones Fox itself was quite happy to make, only with the sort of hedging that makes it hard to call them "false." I mean, how are people supposed to know what they think if you don't help them out a little?