Topic: Media Research Center
It's something of a spectacle to see the Media Research Center -- home of some of the shoddies "media research" -- lash out at the allegedly shoddy media research of others. Taht's what Tim Graham did in a Dec. 4 post:
Washington Post columnist Dana Milbank wrote a column that was so slavishly pro-Biden that it was retweeted by Biden's chief of staff Ron Klain. The preposterous title was "The media treats Biden as badly as — or worse than — Trump. Here’s proof."
As often happens, liberals avoid actually reading articles. That's so old-fashioned, having an actual human read prose. Instead, they assign some cold computer algorithm to make their argument.
[...]
I cry "BS," a mountain of BS. You don't assign a computer to read for adjectives in 200,000 articles and claim media bias. There's no way we at NewsBusters would be able to read 200,000 articles in 15 minutes and say "Voila!" But who needs a human media analyst? Milbank's clown car is speeding down the information superhighway:
We've cried BS over the MRC's shoddy studies, which we've documented as focusing on a small sliver of media (yet somehow never Fox News) to make dubious conclusions about "the media," denies the existence of "neutral" coverage, pretends that subjective concepts such as "spin" and "bias"can be identified objectively, and refusing to make the underlying data public so readers can judge for themselves.The MRC method of handcrafted research by biased researchers works no better -- and, arguably, much worse -- than Milbank's method.
Following Milbank's finding that "Biden’s press for the past four months has been as bad as — and for a time worse than — the coverage Trump received for the same four months of 2020," Graham unironically accused Milbank of being biased:
Naturally, Milbank thinks the "mainstream" media should have a dramatic bias in Biden's favor, since that's in favor of "democracy." Their arrogance in dismissing the entire Republican half of America is perennially breathtaking.
Then again, Graham thinks that all media should have a dramatic bias in favor of Donald Trump and all other Republicans.
Bill D'Agostino followed up with a Dec. 5 post similarly attacking Milbank:
On Sunday’s Reliable Sources, CNN host Brian Stelter entertained the laughable notion that the media have been harder on President Biden than they were on his predecessor, President Trump.
Stelter got this ridiculous idea from a piece by Dana Milbank, who claimed to have analyzed the “tone” of hundreds of thousands of news articles with artificial intelligence. And supposedly, this analysis found that yes, the media were in fact being tougher on Biden:
[...]
Let’s leave aside the notion that harsh coverage of Biden amounts to “the murder of democracy,” and instead talk about Milbank’s conclusion that Biden has been treated at least as harshly as Trump was. To a reasonable person, findings this transparently at odds with reality would suggest there was a problem with either the methodology or the AI the researchers used. As they say in statistics, “garbage in, garbage out.”
D'Agostino, being a loyal MRCer, went on to cite his employer's so-called "research" as if they were legitimate:
As it happens, the Media Research Center has done its own research comparing coverage of Presidents Biden and Trump. Back in April, analysts took an exhaustive look at all evening news coverage from the top three broadcast networks (ABC, CBS, and NBC) and found that during the first three months of each President’s term, the coverage of Biden was 59% positive, while for Trump it was 89% negative.
While Milbank references this three-month “honeymoon” for Biden in his analysis, he does not seem aware that relative to the rest of Trump's presidency, those first three months were also something of a honeymoon for him. Between 2017 and 2021, the coverage of Trump averaged between 90% and 92% negative. Is Milbank suggesting that the current numbers are just as bad for Biden?
Note that the study from last April made no evaluation of right-wing media outlets for comparison purposes, which makes one wonder what the MRC is afraid of. Similarly, the summary of studies purporting to examine "the media" only examined the three networks' evening news -- not a representataive sample of "the media" -- pretended "negative" coverage was determined objectively, again excluded any analysis of Fox News, and refused to release the raw data.
Despite the utter lack of credibility of the MRC's "media research," D'Agostino concluded by pretending otherwise and insisting one didn't need actual research to see the purported bias: "We could go one like this, but there’s no point. Granted, the above research focused on television coverage, but it's not as though the print media were drastically softer on Trump. Milbank’s findings are as much a reflection of reality as a scientific paper concluding that goldfish are more intelligent than dolphins — and that’s obvious to almost everyone. Except to Stelter and [guest Eric] Boehlert, that is."
The MRC wasn't done beating up on Milbank for coming to forbidden conclusions. When a New York Times article referenced Milbank's work, Clay Waters sniped in response in a Dec. 11 post: "Actual humans know Trump’s coverage was far more negative than Biden’s has been, no matter how a computer algorithm may parse coverage." He linked to Graham's post as purported evidence of that.
Graham returned on Dec. 16 to bash Milbank some more:
From the Department of I Told You So: Washington Examiner media reporter Becket Adams discovered that Washington Post columnist Dana Milbank's "study" comparing Trump coverage to Biden coverage using a computer turned out to be hot garbage, despite being quoted and celebrated on CNN and MSNBC.
Or as Adams put it, "after some additional digging, most notably by Don’t Walk, Run! Productions, which did the lion’s share of assessing FiscalNote’s data, it turns out Milbank’s theory is even flimsier than initially suspected."
What Graham left unmentioned: At least Milbank released the raw data of his study -- something Graham's MRC has never done on any of its "media research."
