Earlier this week, I wrote for GBH News about a study showing little support for the core principles of journalism. Joshua Benton of the Nieman Journalism Lab has done an exceptionally deep dive into the numbers and has concluded that they don’t say what the study’s authors claim.
Benton’s explanation is that the Media Insight Project took unambiguous support for certain journalistic verities and watered it down by pairing it with findings that showed a more dubious view of the press. Benton writes:
Its top-line finding — summarized by a [Washington] Post headline writer as “Bad news for journalists: The public doesn’t share our values” — is bogus. Or, at a minimum, unsupported by the methodology in use here. There is no reason to believe, based on this data, that Americans have somehow abandoned the basic values of democratic governance, or that we noble journalists are left to fight the lonely fight for accountability.
But Tom Rosenstiel, executive director of the American Press Institute, one of the organizations that sponsored the study, replies at the Columbia Journalism Review that Benton’s methodology is itself flawed:
Researchers caution against trying to draw conclusions from any one individual item without considering the full set.
We fear this is the mistake Josh has made.
My quick takeaway is that Benton gets the better of the dispute. But read both pieces and see what you think.
5 thoughts on “Fun with numbers, media trust edition”
Does anyone spot-check survey results by conducting thorough interviews of participants to see how well the interpretation of their survey answers reflects what they really seem to believe on the topic?
All the time. Both when I was a fulltime journalism professor and when I was top editor at a B2B broadband magazine. I’ve been self-funding an update of a 2002 survey I did on the synergy between humanitarian aid organizations and the journalists who cover them. I and my colleagues have been testing questions for two years and we’re still not ready to fly the survey. All responsible academics and pollsters do the same.
I appreciate that Benton’s piece had very specific criticisms of the study’s methodology. Rosenstiel’s response about the study questions being interconnected and not easy to derive meaning from individually sounds plausible, but I don’t have the expertise needed to judge this and I’m hoping the dialogue between them continues.
OTOH, Benton’s claim that the paper is “mistrust porn” seems harsh. I can’t see how any thinking person could read that paper and dismiss it so casually.
Benton seems to come at this study solely from the standpoint of a tireless defender of the good name of journalism, vanquishing yet another slanderous attack. In the process, he seems to miss the study’s main point, which is that journalists have an opportunity here. They can use Haidt’s moral foundation framework to better understand differences in values between people so they can better reach their audience and broaden the appeal of their work.
To me, this isn’t an attack on journalists, but more of a “Men are from Mars Women are from Venus” sort of thing. Or maybe “Journalists are from Pluto and some readers are from Neptune.”
I’m not a wild fan of Haidt and I’ve seen a lot of silly (and good) research papers at AEJMC based on the framework — often misused in social media studies (most social media traffic is not “journalism” the way Haidt originally thought).
I’m leaning toward Benton here because I agree with the answers to 18 of the 20 questions and don’t entirely disagree with answers to the other two. I doubt if I can be considered distrustful of journalism in real life. I’ve had a wild career, but mostly in what is defined by my generation as conventional journalism. But I should probably find time in a month or so to ask Tom for the raw data and play around with it.
My preliminary opinion is colored by the squishiness of the 20 questions. I tend toward more specific questions (often with parenthetical examples) and in this case would have probably “doubled” 10 of the 20 questions (that is, asked about the same ideas in different ways). The 30 questions would have gone to a smaller sample of about 2000 (still large) to stay within budget.
I think we need to be careful about saying journalism is in “decline,” by the way.
Local journalism is absolutely in decline by any measure.
National journalism is now highly fragmented… very few huge-circ magazines and must-watch TV.
BUT more documentaries and long-form print journalism (“books”) are being created– up 10-fold in 30 years. That’s where minorities and women find a voice. OMG, there may be more female commentary in the Hemingway 3-parter than in all other Burns documentaries combined!
National media today rarely covers federal agencies on a continuing basis, but B2B does.
The most “powerful” national media organization in circulation and staff is the NYT, which is terrific, but barely covers issues of importance to non-elites in NYC itself. It did not even cover the primary AOC won to gain a lock on the general election. It badly mis-reported Occupy Wall Street and continues to misreport the industry I cover (broadband). But fortunately, Congress believes our reporting on that more than it believes the NYT. Does that mean I don’t trust and cherish what NYT is doing? Silly. Of course I trust and cherish. But not on everything.
Same with Globe. Sent a personal note to reporter who usually does better work, pointed out shoddy reporting on a story, and he flipped me off. Does that mean I don’t cherish the Globe? Also absurd.
Comments are closed.