Does fake news spread faster on Twitter?

You may recall last week a spate of stories and tweets claiming that fake news spreads further and faster on Twitter.  For instance, there is Steve Lohr at the NYT, who doesn’t quite get it right:

And people, the study’s authors also say, prefer false news.

As a result, false news travels faster, farther and deeper through the social network than true news.

That struck me as off-base, and you can find other offenders, so I went back and read the original paper by Vosoughi, Roy, and Aral.  And what did I find?:

1. The data focus solely on “rumor cascades.”  The paper does not establish the relative ratio of fake news to real news, for instance.  The main questions take the form of “within the data set of rumor cascades, what can we say about those cascades?”

2. It still may (or may not) be the case that real news has its major effects through non-cascade mechanisms.  Most people are convinced of 2 + 2 = 4, but probably not because they heard it through a rumor cascade.

3. Within the universe of rumor cascades, this paper measures average effects.  It does not mean that at the margin fake news is more powerful.  For instance, the rumor “Hillary Clinton is Satan” may have been quite powerful, but that does not mean a particular new rumor can achieve the same force.  2 + 2 = 5 won’t get you nearly as far in terms of retweets, I suspect.

4. Overall the results of this paper remind me of another problem/data issue.  At least in the old days, children’s movies used to earn more than films for adults, as stressed by Michael Medved.  That doesn’t mean you have a quick money-making formula by simply making more movies for kids.  It could be a few major kid’s movies, driven perhaps by peer effects, suck up most of the oxygen in the room and dominate the market.  And then, within the universe of cascade-driven movies, kid’s movies will look really strong and indeed be really strong.  That also doesn’t have to mean the kid’s movies have more cultural influence overall, even if they look dominant in the cascade-driven category.  In this analogy, the kid’s movies are like the fake news.

5. I am not sure how much the authors of the paper themselves are at fault for the misunderstandings.  They can defend themselves on the grounds of not being literally incorrect in their statements in the paper.  Still, they do not seem to be going out of their way to correct possible and indeed fairly likely misinterpretations.

6. The strongest argument for the coverage of the authors’ paper is perhaps the coverage of the authors’ paper itself.  The incorrect interpretations of the result did indeed spread faster and further than the correct interpretations.  I even delayed the publication of this post by a few days, if only to make its content less likely to be true.

Comments

One man's "fake news" is another man's priors ...

Speaking of the opinion of others:

"We are experiencing a fundamental paradigm shift in our relationship to knowledge. From the 'information age', we are moving towards the 'reputation age', in which information will have value only if it is already filtered, evaluated and commented upon by others."

Say Goodbye To the Information Age: It's All About Reputation Now

Anti-Straussian. Clarity of endorsement is essential.

Don't forget that Twatter management defines "fake news" as anything that gives the lie to their Marxist agenda.

The replication crisis in the social sciences is so extreme that I think it is often best simply to assume the opposite of whatever you are told. Or to keep in mind your grandmother is always right and your university professor always wrong. So does your Boba agreed with this finding? If so, it is probably true.

This was such a clear effort to justify the FANG efforts at driving the Right off the internet that it was bound to be wrong. But of course as politically convenient Fake News it was also bound to spread fast - but is that proof that they were right?

FANG?

It's a long, sharp tooth. Seems off topic, but some people just talk about whatever they want to talk about, I guess.

Facebook, Amazon, Netflix, and Google (FANG)

Thanks -

Haha! Seriously? You must be f*cking poor. Explains a lot.

Folly has gone democratic. In times past it was the king, the king's court, the ruling class, the elites, which is to say the powerful, who were susceptible to folly, to wooden-headedness to use Barbara Tuchman's term, resulting in a history of humanity punctuated by a succession of calamities. Thanks to social media, however, wooden-headedness and its consequences has gone democratic: no longer is it the king, the king's court, the ruling class, the elites, which is to say the powerful, who are responsible for the calamities, it is us, the people, the voters. Don't blame stupidity on the king, blame stupidity on the wooden-headedness in the voting booth. Alas, voters are no more capable of revelation than kings; indeed, less so since nobody asks voters the difficult questions, and social media provides confirmation of the voters' wisdom, with even greater effect than the sycophants in the king's court. And voters have their scapegoats, whether it be the king or the Russians (who play their dirty tricks - and manipulate - on social media). God gave us minds to think. Don't blame the king or the Russians or social media if we choose not to use them.

To be clear, I believe Mueller's investigation into Russian interference has value, not because the election outcome was the fault of the Russians but because it provides an excuse for the voters' wooden-headedness. Tell voters they are stupid and they will dig in their heels as if to confirm that they are, but tell voters the Russians did it, and at least it's possible that the wooden-headedness will soften a bit.

I think the Russia's to blame narrative is classic Iron Law of Institutions at work.

If the powers that be inside the Democratic Party can make 2016 about something else, they won't ever have to answer for it professionally.

I am more likely to pass a story along that is novel than factual. For instance, https://www.theonion.com/home-depot-releases-new-bluetooth-cordless-hose-1819592894. My family and friends might mute me if I became too pedantic and fastidious by tweeting Encyclopedia Britannica or chemistry/science papers.Thinking at the margins, most people use social media to connect to those around them or those they wish were part of their in group. So we are typically not high brow in the endeavor. The problem comes when the recipient does not intuit that the cordless water sprayer is only possible in the Star Trek universe. Or take dihydrogen monoxide. Think about it... I had acquaintances who were convinced of this hoax. Misinformation is only easy if we are too biddable.

The larger problem comes when misinformation is paired with information. Obscurantism is rampant. We are in a golden age of yellow journalism. Skepticism and deliberation are healthy responses.

I had not heard of the dihydrogen monoxide hoax. I think I need a long stiff glass of it in highly diluted form now. Good grief.

"It is the natural tendency of the ignorant to believe what is not true. In order to overcome that tendency it is not sufficient to exhibit the true; it is also necessary to expose and denounce the false." -H. L. Mencken

Fake news is often designed and optimized to travel fast, hence, it has an advantage over true news.

So wrote Meg McCardle a couple of days ago.

Yep. A bunch of people sit down and make up crap, competing for clicks and shares, sometimes using A/B testing. The ones that make it up to the top get noticed and win the coveted label of “fake news,” and participate in the rumor cascade.

The rest perish, sad, unshared, and unclicked. Not part of the rumor cascade. This is how Buzfeed, Business Insider, and all the other clickbait factories on the web work, too.

The fake news from the MSM works differently. I don’t think it’s intentional. Reporters, when they find an “angle,” get tunnel vision and confirmation bias for the rest of their reporting on the story. The truth is boring, complex, and nuanced, and is never correctly reported. The result is that boring, nuanced, complex truth doesn’t make it to the paper, and all you get is stories with exciting findings that are poor approximations of the truth.

The fake news spreads faster than true news doesn't really tell us anything we didn't already know. Really surprising, exciting or scandalous gossip is spread more quickly and because what makes something surprising/exciting is that its the sort of thing that happens rarely those stories that spread quickly are more likely to be false.

What is interesting in the paper are some of the more specific results. For instance, I found it surprising and interesting that individuals with low numbers of followers seem to be more active in repeating fake stories (perhaps they care less about their reputation in social media contexts?) than those with higher numbers of followers.

Did you see Steve Sailer's post about this?

http://www.unz.com/isteve/that-popular-science-article-on-false-tweets-isnt-really-true/

#3 is a category error. Hillary Clinton is actually Satan, so this is not fake news.

I'm a sceptic. You'd think Satan would be more competent.

I am skeptical of the article itself. This is an ignorant position with respect to the article because I didn't read it. But I do know from first-hand experience that it is extremely difficult to draw inferences like this without controlled experimentation; the confounding variables for why certain content might get more / faster distribution than other are very strong and my instinct is to trust that the standard methodology -- regression -- really controls for them.

To your points:
1) 82,605 were false, 24,409 were true and 19,287 were mixed, corresponding to 1,699 false, 490 true and 259 mixed rumors (see supplemental information, p. 12 at http://science.sciencemag.org/content/suppl/2018/03/07/359.6380.1146.DC1).
2) The study’s interest is in broadly-defined “news” (i.e., “rumors”), not common knowledge like 2+2.
3) Of course it measures average effects. The alternative would be to measure outliers, which say little if anything about trends.
4) Again this gets back to what’s impactful on average, not what’s a sure-fire “hit.” By the Law of Large Numbers, the larger a sample gets, the more closely it approximates the true tendencies of the population.
5) Similarly, I’m not sure how much you at fault for your shallow read of the article, nor for your misunderstanding of how statistics work.
6) Misreporting of the article wasn’t the result of the authors’ misrepresenting it; it was that the media did so. Indeed, the authors have sought to further help readers put their study into perspective since publication: https://twitter.com/dkroy/status/974251282071474177

Comments for this post are closed