Sources: Kevin Roose's tweets about Facebook stories with the highest engagement as per CrowdTangle worried the News Feed team, who considered them inaccurate
What news should people see when they come to Facebook? — In the old days, your answer might have been “whatever they want to see,” or even “who cares?” Tweets: @oliverdarcy , @johnwhegeman , @vivian , @zephyrteachout , @kevinroose , @benedictevans , @mathewi , @emilybell , @gpingersoll , @caseynewton , and @ethanz See also Mediagazer Tweets: Oliver Darcy / @oliverdarcy : @CaseyNewton reports that @kevinroose's tweets showing the top content on Facebook — most often from right-wing pages — have “been driving people at Facebook absolutely crazy.” https://www.getrevue.co/... John Hegeman / @johnwhegeman : @kevinroose 1/6 Some important things to consider about these lists. This isn't a criticism of your analysis, because you're working with accurate engagement data from @CrowdTangle. But these lists don't represent what most people see on FB. Vivian Schiller / @vivian : Fascinating look at Facebook reach vs engagement as measurement of impact from @CaseyNewton. Kudos to @kevinroose who is driving FB towards greater transparency thru his reporting. https://www.getrevue.co/... Zephyr Teachout / @zephyrteachout : Facebook's business model is destroying democracy. Congress should ban it. Amazingly, there are other options: like paying for a service (telephones!). $1 a month and we'd have social media. https://twitter.com/... Kevin Roose / @kevinroose : Reminder that this data is provided by Facebook, collected using Facebook's own data tool, sorted according to that tool's default popularity metric. https://twitter.com/... Benedict Evans / @benedictevans : Thread: the links that get lots of comments on Facebook are not necessarily the ones that most people on Facebook are looking at https://twitter.com/... Mathew Ingram / @mathewi : Appreciate the clarification, but it still shows the content that is most engaged with — the stuff that the largest number of users have interacted with — is a right-wing trash fire. Impressions are largely meaningless, as I'm sure you know https://twitter.com/... Emily Bell / @emilybell : It would be one of the most helpful things Facebook could do in its endless ‘fight against misinformation’ - make available more in-depth attention data on stories circulated on FB, rather than limiting and degrading that access https://twitter.com/... Geoffrey Ingersoll / @gpingersoll : Probably because they leave the public with a totally false impression ... but hell, why let truth get in the way. https://twitter.com/... Casey Newton / @caseynewton : I wrote about @kevinroose's CrowdTangle tweets and why they exasperated Facebook and also convinced them to share some really useful data https://www.getrevue.co/... https://twitter.com/... Ethan Zuckerman / @ethanz : On the one hand, this is a helpful clarification. The lists Kevin Roose posts are disturbing and disconcerting and lead to understandable concerns about FB's affect on civic dialog. On the other hand, if FB will not make more nuanced data available, what else can we conclude? https://twitter.com/... See also Mediagazer Expand More For Next 2 Unexpand More For Next 2
@CaseyNewton reports that @kevinroose's tweets showing the top content on Facebook — most often from right-wing pages — have “been driving people at Facebook absolutely crazy.” https://www.getrevue.co/...
@kevinroose 1/6 Some important things to consider about these lists. This isn't a criticism of your analysis, because you're working with accurate engagement data from @CrowdTangle. But these lists don't represent what most people see on FB.
Fascinating look at Facebook reach vs engagement as measurement of impact from @CaseyNewton. Kudos to @kevinroose who is driving FB towards greater transparency thru his reporting. https://www.getrevue.co/...
Facebook's business model is destroying democracy. Congress should ban it. Amazingly, there are other options: like paying for a service (telephones!). $1 a month and we'd have social media. https://twitter.com/...
Reminder that this data is provided by Facebook, collected using Facebook's own data tool, sorted according to that tool's default popularity metric. https://twitter.com/...
Thread: the links that get lots of comments on Facebook are not necessarily the ones that most people on Facebook are looking at https://twitter.com/...
Appreciate the clarification, but it still shows the content that is most engaged with — the stuff that the largest number of users have interacted with — is a right-wing trash fire. Impressions are largely meaningless, as I'm sure you know https://twitter.com/...
It would be one of the most helpful things Facebook could do in its endless ‘fight against misinformation’ - make available more in-depth attention data on stories circulated on FB, rather than limiting and degrading that access https://twitter.com/...
I wrote about @kevinroose's CrowdTangle tweets and why they exasperated Facebook and also convinced them to share some really useful data https://www.getrevue.co/... https://twitter.com/...
On the one hand, this is a helpful clarification. The lists Kevin Roose posts are disturbing and disconcerting and lead to understandable concerns about FB's affect on civic dialog. On the other hand, if FB will not make more nuanced data available, what else can we conclude? htt…
“As IG was creating new thresholds for suspending or deleting accounts, an internal analysis found that users whose activity suggested they were Black were 50% more likely to have their accounts disabled” The thresholds were tweaked to reduce effects https://www.wsj.com/...
I would also question how reliable using multicultural affinity is because I once checked mine on Facebook and it said African American. Not sure why. https://twitter.com/... https://twitter.com/...
Already excited to read Deepa's piece in 2-3 years about how Facebook execs decided to scrap these studies and bury their findings. https://twitter.com/...
Facebook is aiming to study and address potential racial bias on its core platform and Instagram unit, in a departure from the company's prior reluctance to explore the way its products affect different minority groups https://www.wsj.com/...
Now, per Facebook, the company is talking to outside experts and groups about how to consistently and reliably study race. It won't use its multicultural affinity data. You can read more here: https://www.wsj.com/...
This kind of goldilocks approach that tries to give out accurate information without contradicting and therefore upsetting Donald Trump doesn't work. If anything, it looks like a message endorsing Trump's views. It's critical that Facebook clearly label this post as false. https:…
This warning seems pretty useless—it might even seem that Facebook is endorsing what Trump is saying and providing a path for more information. https://twitter.com/...
FB's labeling of Trump's false and inflammatory post is irresponsible. Rather than call out the falsehood, the label is worded so that some will read it as an endorsement of a false statement. https://twitter.com/...
Facebook began labeling political posts today, but experts are already pointing out flaws. Trump posted blatant lies about a “rigged election.” Biden encouraged people to “vote Donald Trump out.” Both posts got the same treatment. False balance. w/ @donie https://www.cnn.com/...
Good thread. At first, this looked like a landmark move by Facebook against misinformation. But as @donie and others pointed out, Facebook gave the same treatment to a Biden post that contained zero false claims. This raises question about what the labeling actually accomplishes.…
So, rather than noting that this is patently false... Facebook leaves the post up as is, doesn't say whether it's true or false, and then just... adds a link to a page telling Trump supporters how to register to vote. If anything, Facebook adding that to his post helped him. http…
We can't call this a “label” in the sense that it indicates anything about the veracity of a Facebook post. If anything it looks like an “official” stamp of approval/corroboration wherever it appears. https://twitter.com/...
“The-answer-to-bad-speech-is-more- speech” isn't even designed well, Facebook. The “warning” looks like an endorsement & link to more info. I know you're loathe to do anything but feign an interest in “neutral” fact-checking (https://www.cjr.org/...) but your policies are a mess …
Sigh. This doesn't look like a challenge. It looks like an offer to get more of the same. Was any research done on the impact of this message on users? The point is to warn them. Do they feel warned by this? I'll bet not. https://twitter.com/...
Ugh, @Facebook can't you guys get this right? You've got billions to spend on it. Grow up. Do the right thing. Call liars out. Stop the racists. Stop the violence inciters. Stop the @POTUS https://twitter.com/...
This is an interesting point. The way Facebook is doing these labels, could it appear to some that the labels endorse the content in the post? @andymstone @guyro https://twitter.com/... https://twitter.com/...
This is part of the plan Zuckerberg announced a few weeks ago to label posts about the election — without saying if it is true or false. https://www.facebook.com/... https://twitter.com/...
Twitter did nothing about Trump's unfounded claims today about a “corrupt election,” and Facebook did as good as nothing. Both companies claim they are taking misinformation about the election seriously. The latest w/@MarshallCohen https://www.cnn.com/...