Sunday, December 18, 2016

Fact-checking the wrong way, featuring PolitiFact

Let PolitiFact help show you the right way to fact check by avoiding its mistakes

Fake and skewed news do present a problem for society. Having the best possible information allows us to potentially make the best possible decisions. Bad information hampers good decision-making. And the state of public discourse, including the state of the mainstream media, makes it valuable for the average person to develop fact-checking skills.

We found a December 11, 2016 fact check from PolitiFact that will help us learn better how to interpret claims and make use of expert sources.

The interpretation problem

PolitiFact believed it was fact-checking Republican Reince Priebus' claim that there was no report available saying Russia tried to interfere with the 2016 presidential election:


Was Priebus saying there was no "specific report" saying Russia tried to "muddy" the election? Here's how PolitiFact viewed it:
"Let's clear this up. Do you believe -- does the president-elect believe that Russia was trying to muddy up and get involved in the election in 2016?" Meet the Press host Chuck Todd asked on Dec. 11, 2016.

"No. 1, you don't know it. I don't know it," Priebus said. "There's been no conclusive or specific report to say otherwise."

That’s wrong. There is a specific report.

It was made public on Oct. 7, 2016, in the form of a joint statement from the Department of Homeland Security and the Office of the Director of National Intelligence. At the time, the website WikiLeaks was releasing a steady flow of emails stolen from the Democratic National Committee and top Hillary Clinton adviser John Podesta.

"The U.S. Intelligence Community (USIC) is confident that the Russian Government directed the recent compromises of e-mails from U.S. persons and institutions, including from U.S. political organizations," the statement said. "These thefts and disclosures are intended to interfere with the U.S. election process."
Based on the context of Priebus' appearance on "Meet the Press," we think PolitiFact botched its interpretation. NBC's Chuck Todd went back and forth with Priebus for a number of minutes on the nature of the evidence supporting the charge of Russian interference with the 2016 U.S. presidential election. The main topic was recent news reports suggesting Russia interfered with the U.S. election to help Republican candidate Donald Trump. Todd's question after "Let's clear this up" had little chance of clearing up that point. Priebus would not act unreasonably by interpreting Todd's question to refer to interference intended to help the Republican Party.

But the bigger interpretation problem centers on the word "specific." Given the discussion between Todd and Priebus around the epistemological basis for the Russian hacking angle, including "You don't know it and I don't know it" in the immediate context, both the "conclusive" and the "specific" definitions of the word address the nature of the evidence.

"Conclusive" means incontrovertible, not merely featuring a conclusion. "Specific" means including a specific evidence or evidences, and therefore would refer to a report showing evidences, not merely a particular (second definition as opposed to the first) report.

In short, PolitiFact made a poor effort at interpreting Priebus in the most sensible way. Giving conservatives short shrift in the interpretation department occurs routinely at PolitiFact.

Was the report PolitiFact cited incontrovertible? PolitiFact offered no argument to that effect.

Did the report give a clear and detailed description of Russia's attempt to influence the 2016 election? Again, PolitiFact offered no argument to that effect.

PolitiFact's "fact-checking" in this case amounted to playing games with the definitions of words.

The problem of the non-expert expert

PolitiFact routinely cites experts either without investigating or reporting (or both) their partisan leanings. Our current case gives us an example of that, as well as a case of giving the expert a platform to offer a non-expert opinion:
Yoshiko Herrera, a University of Wisconsin political scientist who focuses on Russia, called that letter, "a pretty strong statement." Herrera said Priebus’ comment represents a "disturbing" denial of facts.

"There has been a specific report, and politicians who wish to comment on the issue should read and comment on that report rather than suggest there is no such report or that no work has been done on the topic," Herrera said.
What relevant expertise does a political scientist focused on Russia bring to the evaluation of statements on issues specific to U.S. security? Even taking for granted that the letter Herrera talks about was objectively "a pretty strong statement," Herrera has no obvious qualification that lends weight to her opinion. An expert on international intelligence issues might lend weight to that opinion by expressing it.

The same goes for PolitiFact's second paragraph quoting Herrera. The opinion in this case gains some weight from Herrera's status as a political scientist (the focus on Russia counts as superfluous), but her implicit opinion that Trump made the error she cautions about does not stem from Herrera's field of expertise.

Note to would-be fact checkers: Approach your interviews with experts seeking comments that rely on their field of expertise. Anything else is fluff, and you may end up embarrassing the experts you cite by relying on their expertise for information that does not reflect their expertise.

Was Herrera offering her neutral expert opinion on Trump's comment? We don't see how her comments rely on her expertise. And reason exists to doubt her neutrality.

Source: FEC. Click image for larger view.

Yoshiko Herrera's FEC record of political giving shows her giving exclusively to Democrats, including a modest string of donations to the campaign of Hillary Rodham Clinton.

Did PolitiFact give its readers that information? No.

The wrap

Interpret comments fairly, and make sure you only quote expert sources when their opinion comes from their area of expertise. Don't ask an expert on political science and Russia for an opinion that requires a different area of expertise.

For the sake of transparency, I advocate making interview material available to readers. Did PolitiFact lead Herrera toward the conclusion a "specific report" exists? Or did Herrera offer that conclusion without any leading? An interview transcript allows readers to answer that question.

PolitiFact had announced that it plans to start making interview materials available as a standard practice. Someday? Somewhere over the rainbow?



Afters

Since the time we started this evaluation of PolitiFact's fact check, U.S. intelligence agencies have weighed in with comments hinting that they possess specific evidence showing a Russian government connection to election-related hacking and information leaks. But even these new reports do not contradict Priebus until the reports include the clear and detailed evidence of Russian meddling--from named and reliable sources.

7 comments:

  1. Exactly. A report by an unknown party (or parties) of the (17 member) ODNI that says it "is confident that the Russian Government directed the recent compromises of e-mails from US persons and institutions" with absolutely no evidence to back it up is neither conclusive, nor specific.

    "We believe... could have"

    "we are not now in a position to attribute this activity to the Russian Government".

    It's nothing more than a misdirect. Retribution for Comey's last minute announcement of the Weiner e-mails.

    ReplyDelete
  2. The story from The Hill does not offer any specifics as to how it is known the Russians interfered with the election.

    We conclude that you missed the point of our post, which concerns epistemology (how we can know things are true).

    ReplyDelete
  3. **It is known through intelligence which is still classified but available to the senate, which is what the Graham alluded to on CNN.**

    Do you know the details of the intelligence available to the Senate? Do I?

    That's the point Priebus was making to Chuck Todd.

    If you don't know the specifics of the intelligence then it's an epistemology of trust (Trust Lindsey Graham; why would he lie?).

    ReplyDelete
  4. **That is the way the US government works; it relies on classified intelligence in making foreign policy decisions, and it stands by the intel unless contradictory info disproves it.**

    Often there's no reason to keep the source of the intelligence secret. For example, Colin Powell referenced specific satellite pictures in making the case to the UN that Saddam Hussein's Iraq possessed WMD. He didn't just ask the UN to trust U.S. intel.

    Is there some reason why this intelligence needs to be kept classified? If so, somebody should say so, right?

    **My question is why does the president-elect hold intelligence agencies in such low regard when these people risk their lives in hostile countries gathering information?**

    Our question is why you're asking that question in response to a post that doesn't even mention Trump.

    ReplyDelete
  5. To repeat the point one last time, this post is about PolitiFact's fact-checking and the proper interpretation of Priebus' statement. Your comments have offered nothing that challenges what we've written. You're running off on a tangent. You're welcome to post on-topic, but run off on your tangents someplace else, please.

    ReplyDelete
    Replies
    1. Have you read the post to which you are replying?

      Delete
  6. Cool story, Bro.

    I deleted your comment.

    This is a site about PolitiFact. We have a narrow and specific scope. We've given you more than enough leeway to stay remotely on topic. There are plenty of websites available to post aimless wish-casting and meandering rants. This isn't one of them.

    Anything else you post going forward will be deleted as well.

    ReplyDelete

Thanks to commenters who refuse to honor various requests from the blog administrators, all comments are now moderated. Pseudonymous commenters who do not choose distinctive pseudonyms will not be published, period. No "Anonymous." No "Unknown." Etc.