Noteworthy because it contradicts Fox News’s data from last week. When Fox asked registered voters whom they trusted more to tell them the truth, the Trump administration or reporters who cover the White House, they split 45/42. Less than a week later, when Quinnipiac asked voters whether they trust Trump more or the news media more to tell them the truth, the split was … 37/52. Did Trump’s credibility nosedive in the span of a week or is he right that “any negative polls are fake news”? (Spoiler: He is not right.)

Note the sharply different numbers among independents in the two surveys. Here’s what Fox saw…

…and here’s what Quinnipiac saw:

Indies are +26 for Trump in one poll, -12 for him in the other. Hmmmm. And that’s not the only odd disparity. In Fox’s poll, white voters trust Trump more than the media by a net of 20 points, 54/34, while white college grads are evenly divided at 44/45. In Quinnipiac’s poll, white voters are evenly split between Trump and the media, 45/46, while white college grads trust the media much more than they do him, 37/55. (Nonwhite voters overwhelmingly trust the media more than Trump in both polls, although more so in Quinnipiac.) There’s a gender disparity in the polls too. In Fox’s data, men trust Trump more than they do the media by 12 points while in Quinnipiac’s poll they trust the media more by 10 points. Women trust the media more than Trump in both polls, but the margins are starkly different. In Fox it’s a five-point advantage, in Quinnipiac it’s … 20.

Both polls used registered voters and both used live phone interviews of similarly sized samples roughly a week apart, so go figure why the numbers are so different. It may be that Quinnipiac’s sample skewed much more heavily Democratic than Fox’s, which was 42D/39R. (Quinnipiac didn’t provide a partisan split.) There’s no way to say definitively who’s right, but it’s worth noting that the job approval rating that Quinnipiac found for Trump is conspicuously poor compared to various other polls taken lately. The RCP average has him at 44.7 percent and several surveys, including Fox’s, have him several points higher than that. Quinnipiac has him at 38 percent(!), easily the lowest mark among pollsters being tracked by RCP and the only one to have him under 40. Not only that, but the last Quinnipiac poll conducted at the time of the inauguration had him at 36 percent(!!). That’s the worst number any pollster has recorded since he became president (although Pew, which isn’t tracked by RCP, had him at 39 percent not long ago). For whatever methodological reason, Quinnipiac is piling up polling for Trump that’s consistently and almost uniquely terrible. And it’s not just on questions about the media, either. It’s everything:

Opinions on most of Trump’s personal qualities also are negative, as American voters say:
55 – 40 percent that he is not honest;
55 – 42 percent that he does not have good leadership skills;
53 – 44 percent that he does not care about average Americans;
63 – 33 percent that he is not level-headed;
64 – 32 percent that he is a strong person;
58 – 38 percent that he is intelligent;
60 – 37 percent that he does not share their values.

Meanwhile, Politico’s finding this:

New Politico poll reflects sharply improved "right track" numbers & decent Trump approval rating. Contrast with media/Left apocalyptic tone: pic.twitter.com/TwDxt7ASAJ

— Guy Benson (@guypbenson) February 22, 2017

Hmmm again. With a little effort, you can cook up a theory that the Fox and Quinnipiac polls aren’t inconsistent and that Trump’s trustworthiness vis-a-vis the media simply took a tumble during the period between when the two polls were conducted. Specifically, the Fox poll was finished on February 13th; Mike Flynn resigned that very evening, followed by days of coverage about how he had misled Mike Pence and how Trump himself had known for weeks about it but had said nothing. The Quinnipiac poll wasn’t conducted until February 16th, days later, so maybe the media buzz about Flynn and Trump really did damage perceptions of Trump’s credibility in the interim. (There was also plenty of chatter during that period about Kellyanne Conway’s credibility, but who knows how many Americans pay attention to cable news blather about a presidential advisor.) My guess, though, is that the differences between the polls are more likely attributable to differences in the samples. And Quinnipiac’s samples, for whatever reason, have been ostentatiously anti-Trump over the first month of his term.