What are the most trusted and least trusted "media sources"? The folks at the University of Missouri's Reynolds Journalism Institute's Trusting News project set out to answer that question. Instead of conducting a survey of a random sample of the population—a pricey procedure—it asked 28 "member newsrooms" to sample their audiences. As the Reynolds Institute's "engagement specialist" admits, those volunteering their views were not representative of the general public; as Market Watch pointed out, about 50 percent rated themselves as liberal or very liberal, while only about 20 percent rated themselves as conservative or very conservative.

Given this ideological tilt, some of the results were surprising.

Market Watch highlights as a measure of trustworthiness the difference between the percentages at which respondents cited the named news sources as "trusted" or "not trusted." The top-rated media by this measure were either British (The Economist, Reuters, The Guardian) or government-supported (public television, BBC, NPR). I suspect the British media, less familiar to American media consumers, just doesn't produce many "not trusted" ratings, leaving almost all respondents willing to rate them on the positive side. And the small number of self-described conservatives in the sample helps to account for the high rating of the government-supported outlets.

The nature of the survey helps account for the net positive ratings of the metropolitan newspapers tested. After all, these papers were surveying their own probably dwindling audiences, who presumably tend to be satisfied with what they're getting. And did anyone outside the Kansas City metropolitan area express a view on the trustworthiness of the Kansas City Star?

What's more interesting to me is that of the two national papers tested, this liberal-leaning sample gave a lower trustworthiness rating to the New York Times than to the Wall Street Journal. This suggests the Times has been squandering its historic boast of fairly covering "all the news that's fit to print."

Every newspaper tested got a net positive trustworthiness rating; every television outfit tested was rated negatively. I suspect, again, this is due to higher percentages being willing to rate the TV networks. Fox News gets a more negative response than CNN, MSNBC and the old-line broadcast networks, which is what one would expect given the liberal tilt of the sample.

The lowest trustworthiness ratings were given to internet sources, led by Occupy Democrats (which I'd never heard of), BuzzFeed, Breitbart, and "social media" (whatever that means to respondents), followed by "Trump." (The only other individual tested was "Limbaugh").

Bottom line: the most surprising results from this liberal-tilted survey were the relatively low rating of the New York Times and the apparent lack of any significant negative response, even from the small group of conservatives surveyed, to the left-wing British-based Guardian and BBC.

UPDATE: Joy Mayer, the designer of the Reynolds Institute Trusting News project, has emailed me, taking partial issue with my conclusions in this blogpost. I'm not sure I agree with all her points, but in the interest of fairness I'm happy to publish the body of her email:

Hi, Mr. Barone.

I saw the story you wrote about my Trusting News work, and I would like to clear something up.

You said that the project set out to answer the question of what the most and least trusted media sources were. That's not true. The project set out to understand user-level patterns of trust in news. As part of that, we asked an open-ended question -- name three sources of news you typically trust and three you don't.

People weren't given a list of news brands and asked to rate them. We asked for top and bottom three. So a brand was only represented if it was in someone's top or bottom three. A bunch of people mentioned that they didn't trust Breitbart, for example, or The Huffington Post. Not nearly as many people listed those sites as in their top three most trusted sources. It doesn't mean no one trusts them -- just that they weren't top of mind for most trusted.

We also were completely transparent in our full report about the skewed sample, acknowledging that the sample skewed more liberal, white and trusting than the overall population. We never intended to do definitive work on the habits of average Americans.

The questionnaires served as part of a larger project, in which 28 partner newsrooms worked with me to talk to their own audiences about what they trust. They followed up the questionnaire by interviewing people in their communities, and I'm using what they learned to better understand the nature of trust. I've started recruiting newsrooms willing to help me test strategies designed to build trust. If you'd be interested in working with me, I'd love to talk to you.

I did get a kick out of the comment on your story -- apparently my mugshot gives away that I wear my politics and my trendiness on my sleeve! Ha! It's been awhile since anyone called me trendy.

Thanks for your time.

Joy