New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

N&P: Discuss governments, nations, politics and recent related news here.

Moderators: Alyrium Denryle, Edi, K. A. Pital

User avatar
The Romulan Republic
Emperor's Hand
Posts: 19176
Joined: 2008-10-15 01:37am

New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by The Romulan Republic » 2019-07-03 01:50am

https://www.nbcnews.com/politics/new-st ... p-n1025306
WASHINGTON — President Donald Trump and his allies have long insisted that Russian's 2016 propaganda campaign on social media had no impact on the presidential election.

A new statistical analysis says it may well have.

The study, by researchers at the University of Tennessee, Knoxville, does not prove that Russian interference swung the election to Trump. But it demonstrates that Trump's gains in popularity during the 2016 campaign correlated closely with high levels of social media activity by the Russian trolls and bots of the Internet Research Agency, a key weapon in the Russian attack.

"Our results show that the weeks when Russian trolls were accumulating likes and retweets on Twitter, that activity reliably foreshadowed gains for Trump in the opinion polls," wrote Damian Ruck, the study's lead researcher, in an article explaining his findings.

The study found that every 25,000 re-tweets by accounts connected to the IRA predicted a 1 percent increase in opinion polls for Trump.

In an interview with NBC News, Ruck said the research suggests that Russian trolls helped shift U.S public opinion in Trump's favor. As to whether it affected the outcome of the election: "The answer is that we still don't know, but we can't rule it out."

Given that the election turned on 75,000 votes in Michigan, Wisconsin and Pennsylvania, "it is a prospect that should be taken seriously," Ruck wrote, adding that more study was needed in those swing states.

He points out that 13 percent of voters didn't make their final choice until the last week before the election.

Ruck said the correlation between troll activity and Trump's popularity remained true even when controlling for Trump's own Twitter activity and other variables.

"It turns out that the activity of Russian Twitter trolls was a better predictor of Donald Trump's polling numbers than his own Twitter activity," he wrote.

Ruck was among a group of researchers who won a Defense Department grant this year to study Russian disinformation campaigns in Georgia, Ukraine, and Belarus through March 2024.

The researchers found to their surprise that the Russian propaganda did not drive down Democrat Hillary Clinton's popularity, even though much of it was aimed at discrediting her. One factor may have been that the target audience was mostly right-wing media consumers who already disliked Clinton.

The study noted that Trump's own policy proposals occasionally seemed to dovetail with Russian propaganda.

For example, Russian propaganda exploited the December 2015 shooting in San Bernardino, California "to stoke fears about Muslim immigration," the study says.

Shortly afterward, Trump announced his support of a ban on Muslims entering the United States.

The president's son-in-law and top adviser, Jared Kushner, in April derided the Russian propaganda effort as "a couple of Facebook ads."

In fact, the Internet Research Agency, a troll farm led by an oligarch close to Putin, conducted "a vast, coordinated campaign that was incredibly successful at pushing out and amplifying its messages," according to cybersecurity firm Symantec, which conducted an in-depth analysts of nearly 4,000 Twitter accounts involved in what U.S. intelligence agencies assess was a Russian-government–sponsored propaganda operation designed in part to help Trump get elected president.

The IRA's disinformation campaign was described in the final report by special counsel Robert Mueller, and in his indictment of multiple Russian nationals. U.S intelligence agencies assessed that the social media manipulation was part of a Russian intelligence operation designed to undermine American democracy by exacerbating divisions, hurt Clinton and help Trump. Experts say the U.S. government and American social media companies have yet to develop a strategy to prevent such manipulation from happening again. Just last week, a different cybersecurity firm exposed an Iranian effort to manipulate U.S. social media with fake accounts.

The IRA's basic strategy, Symantec found, was to use a small core of Twitter accounts to push out new content. And they harnessed a wider pool of automated accounts to amplify those messages.

"We also see that 91 percent of first retweeters of IRA tweets were non-IRA bots," the Tennessee study says, "which suggests that propaganda spread into networks of real U.S. citizens."

The Trump administration has not developed a strategy to deal with the social media propaganda that is already cropping up in connection with the 2020 election, and Trump continues to downplay the issue of foreign election interference. Last week, he sparked a furor when he seemed to joke about it during a joint appearance with Russian President Vladimir Putin, saying, in a mocking tone, "Don't meddle in the election."

The White House did not immediately respond to a request for comment.
This is the line that really stood out to me:
"It turns out that the activity of Russian Twitter trolls was a better predictor of Donald Trump's polling numbers than his own Twitter activity"
This also lends weight to Jimmy Carter's recent comments to the effect that Trump's election was not legitimate.

Consider how close the election was. A few tens of thousands of votes in three key swing states. Another one percent support matters in a race like that. Don't tell me it didn't make a difference.
"I know its easy to be defeatist here because nothing has seemingly reigned Trump in so far. But I will say this: every asshole succeeds until finally, they don't. Again, 18 months before he resigned, Nixon had a sky-high approval rating of 67%. Harvey Weinstein was winning Oscars until one day, he definitely wasn't."-John Oliver: https://youtube.com/watch?v=zxT8CM8XntA

User avatar
K. A. Pital
Glamorous Commie
Posts: 20772
Joined: 2003-02-26 11:39am
Location: Elysium

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by K. A. Pital » 2019-07-03 03:43am

If you can impact a US election with only a handful of Facebook ads that cost barely anything, gotta say Russia can rule the world safely lol.
Lì ci sono chiese, macerie, moschee e questure, lì frontiere, prezzi inaccessibile e freddure
Lì paludi, minacce, cecchini coi fucili, documenti, file notturne e clandestini
Qui incontri, lotte, passi sincronizzati, colori, capannelli non autorizzati,
Uccelli migratori, reti, informazioni, piazze di Tutti i like pazze di passioni...

...La tranquillità è importante ma la libertà è tutto!
Assalti Frontali

User avatar
mr friendly guy
The Doctor
Posts: 10522
Joined: 2004-12-12 10:55pm
Location: In a 1960s police telephone box somewhere in Australia

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by mr friendly guy » 2019-07-03 05:32am

The obvious solution to Russian trolls affecting elections, is not to have them . Just like how Trump envisages. :lol:
Never apologise for being a geek, because they won't apologise to you for being an arsehole. John Barrowman - 22 June 2014 Perth Supernova.

Countries I have been to.
Australia, Canada, China, Denmark, Ecuador, Finland, Germany, Malaysia, Netherlands, Norway, Singapore, Sweden, USA.
Always on the lookout for more nice places to visit.

User avatar
EnterpriseSovereign
Jedi Council Member
Posts: 2472
Joined: 2006-05-12 12:19pm
Location: High orbit

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by EnterpriseSovereign » 2019-07-03 12:06pm

mr friendly guy wrote:
2019-07-03 05:32am
The obvious solution to Russian trolls affecting elections, is not to have them . Just like how Trump envisages. :lol:
Not have elections, or not have trolls? :lol:
It's no use debating a moron; they drag you down to their level then beat you with experience.

Just because you have the attention span of a fruit fly doesn't mean the rest of us are so encumbered.

"As you know science is not fact"- HuskerJay
"The Delta Fyler [sic] isn't even a shuttle craft" -HuskerJay69
"The Dominion War wasn't really all that bad"- Admiral Mercury

User avatar
The Romulan Republic
Emperor's Hand
Posts: 19176
Joined: 2008-10-15 01:37am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by The Romulan Republic » 2019-07-03 12:15pm

K. A. Pital wrote:
2019-07-03 03:43am
If you can impact a US election with only a handful of Facebook ads that cost barely anything, gotta say Russia can rule the world safely lol.
"It was just a handful of adds that barely cost anything" line is a favorite of the denialists (because, you know, a little hijacking of our election by a foreign despot is okay). Jared Kushner is quoted in the article using that line almost verbatim.

As opposed to the facts, which are that it was a concerted effort by Russian oligarchs and Russian military intelligence over a period of years, involving Facebook adds, hacking and dissemination of Democratic Party emails, fake identities and web pages, and tens of thousands of social media posts.

But yeah, it costs a lot less money to effect regime change this way than to do it the old-fashioned way, with bombs and tanks (and is also a lot easier to deny, incidentally). And yeah, the Kremlin is taking a pretty good stab at ruling the world for what on the international scale is effectively pocket change. And that should scare the hell out of us.
"I know its easy to be defeatist here because nothing has seemingly reigned Trump in so far. But I will say this: every asshole succeeds until finally, they don't. Again, 18 months before he resigned, Nixon had a sky-high approval rating of 67%. Harvey Weinstein was winning Oscars until one day, he definitely wasn't."-John Oliver: https://youtube.com/watch?v=zxT8CM8XntA

User avatar
K. A. Pital
Glamorous Commie
Posts: 20772
Joined: 2003-02-26 11:39am
Location: Elysium

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by K. A. Pital » 2019-07-03 02:24pm

What is this? I am sorry, let me elaborate.

Of course Russian oligarchs and such, as well as the government, constantly work to further their own goals and to undermine competing world powers. It is the basics of international politics.

But only now suddenly... not just the US can impact world politics? Or its own internal politics? Is that supposed to scare me? Why?
Lì ci sono chiese, macerie, moschee e questure, lì frontiere, prezzi inaccessibile e freddure
Lì paludi, minacce, cecchini coi fucili, documenti, file notturne e clandestini
Qui incontri, lotte, passi sincronizzati, colori, capannelli non autorizzati,
Uccelli migratori, reti, informazioni, piazze di Tutti i like pazze di passioni...

...La tranquillità è importante ma la libertà è tutto!
Assalti Frontali

User avatar
The Romulan Republic
Emperor's Hand
Posts: 19176
Joined: 2008-10-15 01:37am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by The Romulan Republic » 2019-07-03 02:34pm

K. A. Pital wrote:
2019-07-03 02:24pm
What is this? I am sorry, let me elaborate.

Of course Russian oligarchs and such, as well as the government, constantly work to further their own goals and to undermine competing world powers. It is the basics of international politics.

But only now suddenly... not just the US can impact world politics? Or its own internal politics? Is that supposed to scare me? Why?
Ah, as I expected. When you argued "Well, it was just a little interference, so it doesn't matter", what you really meant was "Russia interfering in America's elections is okay because the US does it too, Both Sides, fuck the US."

No, one crime does not justify another. If my neighbor is a murderer, and I break into his house and gun him down, I will still go to prison. Its wrong if the US does it, its wrong if the Kremlin does it. And if you will defend the expansion of fascist power just because it fucks over the US, then you (like most of the "anti-establishment" crowd) will always be very easy for fascists to manipulate.
"I know its easy to be defeatist here because nothing has seemingly reigned Trump in so far. But I will say this: every asshole succeeds until finally, they don't. Again, 18 months before he resigned, Nixon had a sky-high approval rating of 67%. Harvey Weinstein was winning Oscars until one day, he definitely wasn't."-John Oliver: https://youtube.com/watch?v=zxT8CM8XntA

Patroklos
Jedi Council Member
Posts: 2449
Joined: 2009-04-14 11:00am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by Patroklos » 2019-07-03 03:34pm

The Romulan Republic wrote:
2019-07-03 12:15pm
K. A. Pital wrote:
2019-07-03 03:43am
If you can impact a US election with only a handful of Facebook ads that cost barely anything, gotta say Russia can rule the world safely lol.
"It was just a handful of adds that barely cost anything" line is a favorite of the denialists (because, you know, a little hijacking of our election by a foreign despot is okay). Jared Kushner is quoted in the article using that line almost verbatim.

As opposed to the facts, which are that it was a concerted effort by Russian oligarchs and Russian military intelligence over a period of years, involving Facebook adds, hacking and dissemination of Democratic Party emails, fake identities and web pages, and tens of thousands of social media posts.

But yeah, it costs a lot less money to effect regime change this way than to do it the old-fashioned way, with bombs and tanks (and is also a lot easier to deny, incidentally). And yeah, the Kremlin is taking a pretty good stab at ruling the world for what on the international scale is effectively pocket change. And that should scare the hell out of us.
What was the total dollar value and/or percentage of Trump and overall election media spending in 2016? Are these figures known?

User avatar
The Romulan Republic
Emperor's Hand
Posts: 19176
Joined: 2008-10-15 01:37am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by The Romulan Republic » 2019-07-03 04:02pm

Patroklos wrote:
2019-07-03 03:34pm
The Romulan Republic wrote:
2019-07-03 12:15pm
K. A. Pital wrote:
2019-07-03 03:43am
If you can impact a US election with only a handful of Facebook ads that cost barely anything, gotta say Russia can rule the world safely lol.
"It was just a handful of adds that barely cost anything" line is a favorite of the denialists (because, you know, a little hijacking of our election by a foreign despot is okay). Jared Kushner is quoted in the article using that line almost verbatim.

As opposed to the facts, which are that it was a concerted effort by Russian oligarchs and Russian military intelligence over a period of years, involving Facebook adds, hacking and dissemination of Democratic Party emails, fake identities and web pages, and tens of thousands of social media posts.

But yeah, it costs a lot less money to effect regime change this way than to do it the old-fashioned way, with bombs and tanks (and is also a lot easier to deny, incidentally). And yeah, the Kremlin is taking a pretty good stab at ruling the world for what on the international scale is effectively pocket change. And that should scare the hell out of us.
What was the total dollar value and/or percentage of Trump and overall election media spending in 2016? Are these figures known?
That's actually a really good question, as it would help indicate just how much bang Russia was getting for their buck, relatively to other election spending.

Some quick Googling gives me 6.5 billion dollars for the presidential and congressional elections, with 2.4 of that being for the Presidential election alone (including primary contests). Not sure how that breaks down though between different types of funding, how much of it was for Democrats vs Republicans (or third party candidates), and I'd guess that it doesn't include things that weren't officially election spending but may still have influenced public opinion, like non-partisan news coverage.

Edit: For Russian interference, the Internet Research Agency spent 46,000 dollars on Facebook ads prior to the elections (vs 81 million dollars on Facebook spent by the Trump and Clinton campaigns combined):

https://time.com/5573537/mueller-report ... erference/
The message arrived after midnight on Election Day 2016. As news bulletins reported that Hillary Clinton had called Donald Trump to concede in the presidential race, Kirill Dmitriev, a Russian sovereign-wealth fund CEO and pal of Russian President Vladimir Putin, checked his device. “Putin has won,” came the message.

The identity of that note’s sender is among the redacted parts of Special Counsel Robert Mueller’s report. While the document chronicles the sweep of Russia’s interference in the election and provides specific examples of the Trump team’s sometimes-artless interactions with Russian officials, it is silent on a key unanswered question: did those Russian efforts actually swing the outcome of the 2016 contest?

It is perhaps no surprise that Mueller’s report does not attempt to answer this politically charged issue. Quantifying the impact of Russia’s influence operation is extremely difficult, as well as outside the core of the special counsel’s mandate. While the facts marshaled by Mueller’s investigators illustrate the scope of Russia’s influence operation—which included a social-media campaign and the hacking and dissemination of Democratic emails—none of the new revelations threaten the validity of Trump’s election victory.

Mueller’s report builds on previous disclosures that Russia had canvassed U.S. election systems looking for vulnerabilities. In Sept. 2017, the Department of Homeland Security told 21 states that hackers had targeted their voting systems. One Mueller indictment from July 2018 indicated that Russian hackers in June and July of 2016 very likely hacked into an unnamed state and stole information related to 500,000 voters. (State officials in Illinois later acknowledged an intrusion by Russian actors, though they said the number of voters whose personal data had been compromised was 76,000.) The indictment also indicates the hackers poked around in counties in Georgia, Florida and Iowa. A Joint Intelligence Bulletin released last month from the FBI and the Department of Homeland Security revealed the Russians targeted all 50 states.


The Brief Newsletter
Sign up to receive the top stories you need to know right now. View Sample
SIGN UP NOW
There are plenty of signs that Russia shaped how Americans perceived the election. That’s why Kathleen Hall Jamieson, director of the Annenberg Public Policy Center at the University of Pennsylvania, calls the foreign actors “discourse saboteurs” in her latest book, Cyberwar: How Russian Hackers and Trolls Helped Elect a President. The Russians not only injected divisive rhetoric into the public narrative via troll accounts on social media, Jamieson says, but also by stealing troves of email from the Democratic National Committee and top Democrats like Clinton campaign chairman John Podesta and then getting them amplified via mainstream media sources.

Facebook executive Colin Stretch told Congress in October 2017 that about 126 million users on the platform were exposed to Russian propaganda masquerading as real news. At the same time, Stretch added, the propaganda was hardly ubiquitous; suspected Russian content amounted to about one out of every 23,000 posts.

Russia’s Internet Research Agency spent just $46,000 on Facebook ads before the 2016 elections, compared to the $81 million the Trump and Clinton campaigns spent on the platform combined. But the raw number may not reflect the operation’s reach. In 2016, Russian agents posted just under 30,000 times on Facebook, yet the operation generated almost 13 million shares, 15 million likes and 1.3 million comments, according to a research team at Oxford University.

Russia ramped up the social-media operation at critical moments, too. There were clear surges in Russian content on Facebook as the primaries and caucuses were getting underway; the numbers dropped off in late spring and early summer when the nominees were decided, only to spike in the final march from the nominating conventions into Election Day.

While it’s virtually impossible to determine how many votes these efforts swayed, the numbers needn’t be huge to be important. In Florida, Trump prevailed by 119,770 votes, or 1.3 percentage points. In North Carolina, Trump’s margin was 173,315 votes, or 3.6 percentage points. Trump wrested Michigan, Pennsylvania and Wisconsin into the Republican column by a margin of 77,744 combined votes. According to a July indictment from Mueller, the Russians knew exactly who the Clinton campaign needed to reach to win the latter Democratic strongholds because the Russians had her strategy memos. Russian-backed trolls targeted 36-social media campaigns at Michigan voters, 25 in Pennsylvania and 54 in Wisconsin, according to data provided to the House intelligence panel.

Still, it’s impossible to chalk up the election result entirely to Russia’s influence—or to posit that the Kremlin’s efforts were a complete success. Consider Russia’s plan to depress African-American turnout. The Oxford study found that of the five top Russian-backed accounts on Twitter, four focused on reaching African-Americans to anger them about Clinton’s record. In 2012, 16% of voters in Michigan were black, according to exit polls. Four years later, the number of 15%. In Wisconsin, the share of the black vote held steady at 7%. It’s true that black turnout declined nationwide in 2016 for the first time in 20 years; the net drop of 7 percentage points from 2012 is the largest on record, according to Pew. But it’s tough to see that as decisive in the trio of states that lifted Trump to victory.

As with any election, it’s impossible to single out any lone factor as the reason why, in the privacy of a polling booth, someone cast a ballot for—or against— an individual. No one can prove Clinton would have won had the Russians not meddled. Mueller’s report illustrates how Russia engaged in an effective campaign to help Trump win the White House. But it doesn’t argue those efforts were decisive.

With reporting by John Walcott/Washington
Of course, the IRA's Facebook ads were only a subset of the Russian campaign, but this does suggest, especially in conjunction with the other study cited in the OP, that Russia was getting a lot more bang for its buck than domestic campaign spending, by carefully targeting and timing their content for maximum effect, and generating large numbers of shares from Americans who took the bait.
"I know its easy to be defeatist here because nothing has seemingly reigned Trump in so far. But I will say this: every asshole succeeds until finally, they don't. Again, 18 months before he resigned, Nixon had a sky-high approval rating of 67%. Harvey Weinstein was winning Oscars until one day, he definitely wasn't."-John Oliver: https://youtube.com/watch?v=zxT8CM8XntA

Patroklos
Jedi Council Member
Posts: 2449
Joined: 2009-04-14 11:00am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by Patroklos » 2019-07-03 05:54pm

https://www.npr.org/2018/02/17/58669034 ... on-warfare

It appears the max claimed, but unproven, monthly expenditure was 1.25 million. So somewhere between .01 and .001 of your 6.5 billion. Or less than the yearly salary of a US infantry platoon (without an operating budget) to put it in government expenditure context. Yawn.

Is it your position that this anemic expenditure of resrourses was enough to out PR the thousands of high priced professional influencer hacks employed by all parties, able to act overtly with thousands of times more money at their disposal?

User avatar
The Romulan Republic
Emperor's Hand
Posts: 19176
Joined: 2008-10-15 01:37am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by The Romulan Republic » 2019-07-03 06:42pm

Patroklos wrote:
2019-07-03 05:54pm
https://www.npr.org/2018/02/17/58669034 ... on-warfare

It appears the max claimed, but unproven, monthly expenditure was 1.25 million. So somewhere between .01 and .001 of your 6.5 billion. Or less than the yearly salary of a US infantry platoon (without an operating budget) to put it in government expenditure context. Yawn.

Is it your position that this anemic expenditure of resrourses was enough to out PR the thousands of high priced professional influencer hacks employed by all parties, able to act overtly with thousands of times more money at their disposal?
By itself? No. But in a close race, could it have made the difference, especially once those US campaigners and their supporters picked up on the Russia propaganda and (knowingly or unknowingly) started re-posting it? Yeah, I think so, and the statistical evidence, while not definitive proof, suggests that that is the case.

And the cost of an infantry platoon isn't really a relevant comparison, because these are two very different kinds of warfare. Yes, disseminating lies to hundreds of millions of people on the internet costs less than arming conventional military forces. That's part of the power of this technology, and the danger it poses. That's how viral media and memes work- there are Youtube channels run by random schmucks who probably struggle to make ends meet that have millions of followers. Does it seem incredible to you that an organization run by powerful oligarchs, or a state intelligence agency, in the service of a propagandist as skilled as Putin (skilled enough he has many "anti-establishment" Leftists fooled into thinking he's their friend and repeating his narratives) could do better? And before you say "well, its just internet shitposting, it doesn't matter"- yeah, actually, it does. What people say influences what people think, and what they do. Speech is a powerful weapon, especially when its disseminated through mass media. Why do you think governments through the ages have made censoring it, and disseminating their own propaganda, a priority?

But at the end of the day, this isn't about what I say or think, or what you say or think, or what anyone else says or thinks- its about what the evidence says. And the evidence says that yeah, it probably did matter. You may not be comfortable with acknowledging that evidence, for any number of reasons- because you don't like to acknowledge that a foreign adversary played the American electoral system like chumps, or because you support Trump and don't want to admit his legitimacy as President is questionable, or because "the other side" says it and you don't want to give them the satisfaction of admitting they were right, or any number of other reasons. But the facts don't go away just because you're uncomfortable with them. And to my mind, a true patriot acknowledges when something is wrong with their country, so that they can address it, rather than pretending that the problem isn't there.

Or maybe you know damn well what happened, and you just don't care.
"I know its easy to be defeatist here because nothing has seemingly reigned Trump in so far. But I will say this: every asshole succeeds until finally, they don't. Again, 18 months before he resigned, Nixon had a sky-high approval rating of 67%. Harvey Weinstein was winning Oscars until one day, he definitely wasn't."-John Oliver: https://youtube.com/watch?v=zxT8CM8XntA

User avatar
mr friendly guy
The Doctor
Posts: 10522
Joined: 2004-12-12 10:55pm
Location: In a 1960s police telephone box somewhere in Australia

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by mr friendly guy » 2019-07-03 07:35pm

EnterpriseSovereign wrote:
2019-07-03 12:06pm
mr friendly guy wrote:
2019-07-03 05:32am
The obvious solution to Russian trolls affecting elections, is not to have them . Just like how Trump envisages. :lol:
Not have elections, or not have trolls? :lol:
Touche, touche.

On an unrelated note, does this mean people who subscribe to this hypothesis agree in principle that countries who banned Facebook are justified because another geopolitical actor could use it to interfere with their internal affairs.
Never apologise for being a geek, because they won't apologise to you for being an arsehole. John Barrowman - 22 June 2014 Perth Supernova.

Countries I have been to.
Australia, Canada, China, Denmark, Ecuador, Finland, Germany, Malaysia, Netherlands, Norway, Singapore, Sweden, USA.
Always on the lookout for more nice places to visit.

User avatar
The Romulan Republic
Emperor's Hand
Posts: 19176
Joined: 2008-10-15 01:37am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by The Romulan Republic » 2019-07-03 07:40pm

mr friendly guy wrote:
2019-07-03 07:35pm
EnterpriseSovereign wrote:
2019-07-03 12:06pm
mr friendly guy wrote:
2019-07-03 05:32am
The obvious solution to Russian trolls affecting elections, is not to have them . Just like how Trump envisages. :lol:
Not have elections, or not have trolls? :lol:
Touche, touche.

On an unrelated note, does this mean people who subscribe to this hypothesis agree in principle that countries who banned Facebook are justified because another geopolitical actor could use it to interfere with their internal affairs.
No, I personally wouldn't say banning Facebook is justified.

More tightly regulating it (as all businesses in America should be more tightly regulated) to ensure that it is not being used (knowingly or unknowingly) to violate Federal election law, and possibly prosecuting those who post under fake identities for fraud, yes.
"I know its easy to be defeatist here because nothing has seemingly reigned Trump in so far. But I will say this: every asshole succeeds until finally, they don't. Again, 18 months before he resigned, Nixon had a sky-high approval rating of 67%. Harvey Weinstein was winning Oscars until one day, he definitely wasn't."-John Oliver: https://youtube.com/watch?v=zxT8CM8XntA

User avatar
mr friendly guy
The Doctor
Posts: 10522
Joined: 2004-12-12 10:55pm
Location: In a 1960s police telephone box somewhere in Australia

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by mr friendly guy » 2019-07-03 10:22pm

The Romulan Republic wrote:
2019-07-03 07:40pm

No, I personally wouldn't say banning Facebook is justified.

More tightly regulating it (as all businesses in America should be more tightly regulated) to ensure that it is not being used (knowingly or unknowingly) to violate Federal election law, and possibly prosecuting those who post under fake identities for fraud, yes.
Note I am referring to countries which ban Facebook, so obvious the US isn't one of them and hence I am not referring to the US when I asked my question. So talking about tightening regulation in the US is kind of irrelevant.
https://en.wikipedia.org/wiki/Censorship_of_Facebook

If country X said to Facebook, some actor whose motives are at odds with our internal affairs, is using Facebook to attempt to undermine our rules, and Facebook refuses to block those posts or were unable to, then is it ok to ban it? Remember how people were saying social media is soooo awesome, during the Arab Spring because it helps people organise to protests various governments.

The way I see it, if the Left (and yes its the Left), now accept that a Facebook adverts counts as interference in a country's internal affairs, then countries who banned Facebook on the same grounds would be perfectly justified if Facebook was unable to comply with regulations to prevent this. I am using the same logic people are who complain about Russian interference. In fact, people who complained about those countries suppressing FREEDOMTM should eat their words, that these countries were right to do so. But I suspect they won't. Because when social media is used to help our side, its democracy in action, its people expressing their views, when it counters our view, its interfering with our internal affairs.
Never apologise for being a geek, because they won't apologise to you for being an arsehole. John Barrowman - 22 June 2014 Perth Supernova.

Countries I have been to.
Australia, Canada, China, Denmark, Ecuador, Finland, Germany, Malaysia, Netherlands, Norway, Singapore, Sweden, USA.
Always on the lookout for more nice places to visit.

Ralin
Sith Devotee
Posts: 2686
Joined: 2008-08-28 04:23am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by Ralin » 2019-07-04 01:28am

The Romulan Republic wrote:
2019-07-03 07:40pm

No, I personally wouldn't say banning Facebook is justified.

More tightly regulating it (as all businesses in America should be more tightly regulated) to ensure that it is not being used (knowingly or unknowingly) to violate Federal election law, and possibly prosecuting those who post under fake identities for fraud, yes.
How exactly would you suggest that China or Syria go about 'tightly regulating' an American company?

User avatar
The Romulan Republic
Emperor's Hand
Posts: 19176
Joined: 2008-10-15 01:37am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by The Romulan Republic » 2019-07-04 02:47am

Ralin wrote:
2019-07-04 01:28am
The Romulan Republic wrote:
2019-07-03 07:40pm

No, I personally wouldn't say banning Facebook is justified.

More tightly regulating it (as all businesses in America should be more tightly regulated) to ensure that it is not being used (knowingly or unknowingly) to violate Federal election law, and possibly prosecuting those who post under fake identities for fraud, yes.
How exactly would you suggest that China or Syria go about 'tightly regulating' an American company?
Same as you regulate any other company: have a set of regulations that company must adhere to to operate in that country, and then fine them for violations.

Now, this is not me condoning China censoring content its government disagrees with. There is a difference between regulating fraudulent behaviour, and censoring dissent (though the unscrupulous can use the one as a pretext for the other).

Just pointing that out because I know perfectly well that's the intended trap here, the intent of the question: to either force me (or others who share my views) to "admit" that Chinese censorship is justified, or else "expose" myself as a hypocrite for supporting American censorship while condemning Chinese censorship. I know that both because its obvious, and because practically every other thing you and mr friendly guy post in the News and Politics forum is a defense of the Chinese dictatorship/an attack on its opponents. I'm not particularly interested in re-arguing that point here, as it is off-topic and will only end in a shit show- I just wanted you to know that I recognize the ploy, and that I am not taking the bait.
"I know its easy to be defeatist here because nothing has seemingly reigned Trump in so far. But I will say this: every asshole succeeds until finally, they don't. Again, 18 months before he resigned, Nixon had a sky-high approval rating of 67%. Harvey Weinstein was winning Oscars until one day, he definitely wasn't."-John Oliver: https://youtube.com/watch?v=zxT8CM8XntA

User avatar
mr friendly guy
The Doctor
Posts: 10522
Joined: 2004-12-12 10:55pm
Location: In a 1960s police telephone box somewhere in Australia

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by mr friendly guy » 2019-07-04 07:19am

The Romulan Republic wrote:
2019-07-04 02:47am
Same as you regulate any other company: have a set of regulations that company must adhere to to operate in that country, and then fine them for violations.

Now, this is not me condoning China censoring content its government disagrees with. There is a difference between regulating fraudulent behaviour, and censoring dissent (though the unscrupulous can use the one as a pretext for the other).

Just pointing that out because I know perfectly well that's the intended trap here, the intent of the question: to either force me (or others who share my views) to "admit" that Chinese censorship is justified, or else "expose" myself as a hypocrite for supporting American censorship while condemning Chinese censorship. I know that both because its obvious, and because practically every other thing you and mr friendly guy post in the News and Politics forum is a defense of the Chinese dictatorship/an attack on its opponents. I'm not particularly interested in re-arguing that point here, as it is off-topic and will only end in a shit show- I just wanted you to know that I recognize the ploy, and that I am not taking the bait.
Ah yes, everytime someone points out Western double standards we are apologists for evil dictatorships wah wah wah. Why don't you just say "why do you love Saddam?" :lol: Let me make it easy for you.

I don't think China should censor Facebook because Russia or the US might put their silly adverts on saying Xi Jinping sucks or something. They justify the ban based on other reasons which I know you'll never bother to research. I will also point, even if you think their original reason is bogus, they can just turn around and change the reason to, "we don't want Russia or US putting Facebook adds which will be detrimental to our internal affairs," AND, here is the important bit, Leftist would be hypocritical to object to that, after putting on such a dog and pony show about Russia interference. Since you obviously know me so well, you should realise that I really love pointing out the hypocrisy of people. And unlike some people, I don't confine it to the Right wing.
Never apologise for being a geek, because they won't apologise to you for being an arsehole. John Barrowman - 22 June 2014 Perth Supernova.

Countries I have been to.
Australia, Canada, China, Denmark, Ecuador, Finland, Germany, Malaysia, Netherlands, Norway, Singapore, Sweden, USA.
Always on the lookout for more nice places to visit.

User avatar
The Romulan Republic
Emperor's Hand
Posts: 19176
Joined: 2008-10-15 01:37am

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by The Romulan Republic » 2019-07-04 08:00am

mr friendly guy wrote:
2019-07-04 07:19am
The Romulan Republic wrote:
2019-07-04 02:47am
Same as you regulate any other company: have a set of regulations that company must adhere to to operate in that country, and then fine them for violations.

Now, this is not me condoning China censoring content its government disagrees with. There is a difference between regulating fraudulent behaviour, and censoring dissent (though the unscrupulous can use the one as a pretext for the other).

Just pointing that out because I know perfectly well that's the intended trap here, the intent of the question: to either force me (or others who share my views) to "admit" that Chinese censorship is justified, or else "expose" myself as a hypocrite for supporting American censorship while condemning Chinese censorship. I know that both because its obvious, and because practically every other thing you and mr friendly guy post in the News and Politics forum is a defense of the Chinese dictatorship/an attack on its opponents. I'm not particularly interested in re-arguing that point here, as it is off-topic and will only end in a shit show- I just wanted you to know that I recognize the ploy, and that I am not taking the bait.
Ah yes, everytime someone points out Western double standards we are apologists for evil dictatorships wah wah wah. Why don't you just say "why do you love Saddam?" :lol: Let me make it easy for you.

I don't think China should censor Facebook because Russia or the US might put their silly adverts on saying Xi Jinping sucks or something. They justify the ban based on other reasons which I know you'll never bother to research. I will also point, even if you think their original reason is bogus, they can just turn around and change the reason to, "we don't want Russia or US putting Facebook adds which will be detrimental to our internal affairs," AND, here is the important bit, Leftist would be hypocritical to object to that, after putting on such a dog and pony show about Russia interference. Since you obviously know me so well, you should realise that I really love pointing out the hypocrisy of people. And unlike some people, I don't confine it to the Right wing.
Yes, I'm sure the Chinese government has some very nice-sounding official reasons for their censorship. But please, do explain to me the necessary, totally-justified reasons behind, say, Marvel's Doctor Strange having to whitewash the character of the Ancient One rather than show a positive portrayal of a Tibetan person in order to avoid Chinese government censorship. Or, far more seriously, the effort to completely erase all mention of Tiananmen Square from Chinese history.

As for your attempts to turn yet another News and Politics thread into an attack on a strawman of my character and values, rather than my actual arguments, I refuse to engage with it. I am not the topic of this thread. Suffice to say that I have never confined my criticisms only to the Right. I have also never supported the invasion of Iraq, contrary to what some on this board have persistently claimed, so leave your Saddam strawman out of it. It is you who whole-heartedly defends tyranny if it happens to march under a certain country's flag, not I. And then you have the gall to try to turn it around and accuse me of doing so to hide your own hypocrisy. And it works, because as we all know (because its been endlessly drilled into the popular consciousness as a self-evident fact) Both Sides Are Just As Bad (but really we mean that the West is worse and anything its enemies do is okay, because the West is bad).

And no, "the Left" is not hypocritical for opposing Chinese censorship while also opposing Russian intervention. What a wonderful argument: "You must accept Russian interference to destroy your democracy. Or you must accept Chinese censorship. Or you are a hypocrite, and therefore have no right to criticize dictatorship." How very convenient for the advocates of tyranny: Every possible answer requires us to accept dictatorship. Of course, that's actually so much bullshit, because there is absolutely no contradiction whatsoever between opposing foreign election interference and opposing state censorship of social media for political purposes, but since I refused to walk into my trap, you'll of course just pretend that I did and call me a hypocrite anyway.

You Useful Idiots are so predictable. Get a new tune. This one has worn very thin.
"I know its easy to be defeatist here because nothing has seemingly reigned Trump in so far. But I will say this: every asshole succeeds until finally, they don't. Again, 18 months before he resigned, Nixon had a sky-high approval rating of 67%. Harvey Weinstein was winning Oscars until one day, he definitely wasn't."-John Oliver: https://youtube.com/watch?v=zxT8CM8XntA

User avatar
Gandalf
SD.net White Wizard
Posts: 15224
Joined: 2002-09-16 11:13pm
Location: A video store in Australia

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by Gandalf » 2019-07-04 09:03am

The Romulan Republic wrote:
2019-07-03 02:34pm
Ah, as I expected. When you argued "Well, it was just a little interference, so it doesn't matter", what you really meant was "Russia interfering in America's elections is okay because the US does it too, Both Sides, fuck the US."

No, one crime does not justify another. If my neighbor is a murderer, and I break into his house and gun him down, I will still go to prison. Its wrong if the US does it, its wrong if the Kremlin does it. And if you will defend the expansion of fascist power just because it fucks over the US, then you (like most of the "anti-establishment" crowd) will always be very easy for fascists to manipulate.
One crime may not justify another, but it makes it hard to care if the US' election was influenced, given the US' overt political interference all over the world.
"Oh no, oh yeah, tell me how can it be so fair
That we dying younger hiding from the police man over there
Just for breathing in the air they wanna leave me in the chair
Electric shocking body rocking beat streeting me to death"

- A.B. Original, Report to the Mist

"I think it’s the duty of the comedian to find out where the line is drawn and cross it deliberately."
- George Carlin

User avatar
mr friendly guy
The Doctor
Posts: 10522
Joined: 2004-12-12 10:55pm
Location: In a 1960s police telephone box somewhere in Australia

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by mr friendly guy » 2019-07-04 09:42am

The Romulan Republic wrote:
2019-07-04 08:00am
Yes, I'm sure the Chinese government has some very nice-sounding official reasons for their censorship.
Yeah, they do actually, but like I said, you're not going to research it anyway. But I can assure you, it sounds more serious than Russia might put some adverts on Facebook. :roll:
The Romulan Republic wrote:
2019-07-04 08:00am
But please, do explain to me the necessary, totally-justified reasons behind, say, Marvel's Doctor Strange having to whitewash the character of the Ancient One rather than show a positive portrayal of a Tibetan person in order to avoid Chinese government censorship. Or, far more seriously, the effort to completely erase all mention of Tiananmen Square from Chinese history.
China didn't tell Marvel to whitewash their character idiot. Hollywood does that by itself. I pointed that out when in the Dr Strange thread when it was coming out, China has entire movies starring Tibetans and they don't give a shit and that Tibetan actors have been feted by Chinese viewers. Not that you or Disney know.

But lets go on, because your example betrays your blatant hypocrisy. Lets say Marvel hires hmm, Richard Gere. :lol: Someone who China hates. China is under no obligation to allow that Marvel film in. That's freedom to choose what films to allow in their country. Its not fucking censorship. Marvel can then choose to change the film to be more likely to win the Chinese market, or they don't have to.

By that fucking logic, if a station refuses to show reruns of the Cosby show in favour of a show more people would want to watch, then its censorship by your logic.
As for your attempts to turn yet another News and Politics thread into an attack on a strawman of my character and values, rather than my actual arguments, I refuse to engage with it.
Paranoia getting to you is it. :D
I am not the topic of this thread. Suffice to say that I have never confined my criticisms only to the Right. I have also never supported the invasion of Iraq, contrary to what some on this board have persistently claimed, so leave your Saddam strawman out of it.
Hint - I accused you of using the same smear the right wing use, not that you support the same thing they do. Jesus fucking Christ.
It is you who whole-heartedly defends tyranny if it happens to march under a certain country's flag, not I.
Ah, the why do you love Saddam. Again. :banghead: But just for the record, I have never accused you of supporting tyranny. I just think you're laughably ignorant about what the US has done and continues to do right to the present day. Given how you perceive what Russia is supposed to have done with Facebook ads being worse than what the US is doing right now in terms of interfering in someone's domestic affairs.
And then you have the gall to try to turn it around and accuse me of doing so to hide your own hypocrisy.
No, I am pretty sure I accused you of using the same smear.
And it works, because as we all know (because its been endlessly drilled into the popular consciousness as a self-evident fact) Both Sides Are Just As Bad (but really we mean that the West is worse and anything its enemies do is okay, because the West is bad).
No, on this particular issue ie interference in another country's internal workings, the US is much worse than Russian Facebook ads. That's just an objective fact. If you tried comparing Russian actions in Ukraine with the US, it would be a more equivalent comparison.
And no, "the Left" is not hypocritical for opposing Chinese censorship while also opposing Russian intervention.
Oh please. The Left thinks interference in their country is bad, but totally ok for the US to do it to another country. Rachael Maddow feels sorry for John "warmonger" Bolton, because Trump suddenly flip flopped and doesn't want to put in the effort to overthrow Venezeuala's government anymore. That's hypocrisy right there.
What a wonderful argument: "You must accept Russian interference to destroy your democracy. Or you must accept Chinese censorship.
Actually its more like, we should all refrain from interference except under extreme circumstances. One reason is, when we do it, we open a Pandora's box and invite someone else to do the same to us. And we know how you feel about someone else doing it to us right. :D
Or you are a hypocrite, and therefore have no right to criticize dictatorship.
More like you're a hypocrite, and a specific argument you made falls apart due to self contradiction. It doesn't preclude you from making another argument against a country you don't like, nor does it imply that other argument would be bullshit. Seriously. That's all you got? A poorly thought out strawman?
" How very convenient for the advocates of tyranny: Every possible answer requires us to accept dictatorship. Of course, that's actually so much bullshit, because there is absolutely no contradiction whatsoever between opposing foreign election interference and opposing state censorship of social media for political purposes, but since I refused to walk into my trap, you'll of course just pretend that I did and call me a hypocrite anyway.
In case you missed it the first time and the second time, what is to stop these states who you accuse of over the top censorship, of suddenly turning around and saying, well my previous reasons for banning Facebook suck, but I got a good one now. The exact same reason you're using. :lol: Facebook adds which interfere in our political workings. Oh wait... these countries aren't democracies, so its totally ok.
You Useful Idiots are so predictable. Get a new tune. This one has worn very thin.
Buddy, I am not the one who is ignorant of the US doing worse. You are like the child who complains about Bob insulting Jim, without realising Jim has for years insulted Bob and several other kids.
Never apologise for being a geek, because they won't apologise to you for being an arsehole. John Barrowman - 22 June 2014 Perth Supernova.

Countries I have been to.
Australia, Canada, China, Denmark, Ecuador, Finland, Germany, Malaysia, Netherlands, Norway, Singapore, Sweden, USA.
Always on the lookout for more nice places to visit.

User avatar
Ziggy Stardust
Sith Devotee
Posts: 2948
Joined: 2006-09-10 10:16pm
Location: Research Triangle, NC

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by Ziggy Stardust » 2019-07-04 11:22am

I'm going to ignore this petty squabbling to get back to the OP a little.

Now, unfortunately, before I even say anything, I feel it necessary to put my disclaimer at the top. I am NOT trying to be an apologist for Russian interference. Since I know people on this forum have a difficult time responding to arguments without trying to appeal to the political motivations of the poster, I feel I have to point it out. I am ONLY commenting on the study mentioned in the OP from a methodological/analytical viewpoint, to try and better understand the inference it gives us. I have 0 overt political motivations for doing this, with the obvious caveat that my perspective is inherently biased by my own left-wing beliefs. I will also point out, for those of you not aware, I am a professional statistician, so I do have the qualifications to comment on these methods.

(This is quite long, there is a TL;DR version at the end).

Anyway, the elevator pitch for the study is essentially, "Trump's gains in popularity during the 2016 campaign correlated closely with high levels of social media activity by the Russian trolls." More specifically, "The study found that every 25,000 re-tweets by accounts connected to the IRA predicted a 1 percent increase in opinion polls for Trump." (Both these quotes from the article in the OP).

If you look at the original paper for the study, we can dig into the results a little more. For example, we can see the results of one of their primary statistical analyses in Figure 3. This test shows the results of "Granger causality tests" from a "Vector Auto-Regression (VAR)". Now, a VAR can be thought of rather simply as a multivariate time series. Though the devil is in the details, conceptually a VAR isn't much different than a simple linear regression of the kind you learn in algebra (remember y=ax+b?). The "Granger causality test" has a rather grandiose name, but it's just a fancy name for a hypothesis test of a set of variables in the regression model. For example, if your model is y=ax+bz+c, a "Granger causality test" on z would simply be testing that b=0 (the reality is slightly more complicated because in the actual time series case you are dealing with sets of variables, some of which are "lagged" in time, but the intuition is the same). Another way to put it for that same toy model would be to say, "Which of the following equations does the best job of explaining the variation in Y: is it y=ax+c or y=ax+bz+c?" This tells us the degree to which including bz improves the overall explanatory power of the model. In the case of this paper, "Y" is a set of polling results, and "bz" are the social media activities of the Russian bots one week before the polling results. Note that, despite the name, this is not actually a test for "causality", it is (like any regression) a test of conditional correlations. For example, it is unable to properly distinguish between X causing Y, Z causing both X and Y, or Z is caused by both X and Y (again, like any regression).

With that explanation out of the way, let's look at Figure 3 again. It shows us that the Granger causality test found a significant increase in the model's explanatory power for explaining Trump's poll results when Russian bot activity one week before is included in the model. It also finds that there is NO significant relationship between Russian bot activity and Clinton's poll results. This is already something that should raise your eyebrows. If the Russian bots were targeting Hillary to help Trump, why do we not see a significant decrease in her polling numbers as a result of bot activity? We only see an increase in Trump's polling numbers. While polls are technically not a zero-sum game, you expect some degree of symmetry in any changes in polling numbers. Someone only goes up if someone else is going down. But this shows Clinton is not going down. Given the time-frame of the data they are using (2015-2016), you EXPECT Clinton's numbers to remain pretty steady (since she was the favorite to be the Democratic candidate from the start) and Trump's to rise (since he emerged rather surprisingly out of a crowded field of Republican nominees). Unfortunately, the authors do not state whether they controlled for this or how. From a strictly mathematical standpoint this doesn't pose a problem for the Granger causality test, but from an inferential standpoint it also hints at the possibility that the causal model is more complicated then X causes Y, and almost certainly involves another set of variables Z that may be common causes of both.

In fact, digging through their methods, it doesn't appear that the authors controlled for ANY covariates in their model. As the authors themselves note in their abstract, "We find that the release of these tweets parallel significant political events of 2016 and that approximately every 25,000 additional IRA re-tweets predicted a one percent increase in election opinion polls for one candidate." Read that agian, "the release of these tweets parallel significant political events of 2016". They go into some detail in the paper showing that there is a correlation between the occurrence of "significant political events" and increases in Russian troll activity. What they DON'T mention, to their great discredit, is that you would also expect a correlation between the occurrence of significant political events and candidate's polling numbers even in the ABSENCE of Russian trolls. The way their model is currently constructed, they are assuming by default that the ONLY mechanism through which significant political events could impact the candidate's polling numbers is through Russian troll activity on Twitter! It's an unavoidable consequence of their model construction, though they don't explicitly state that. I think on its face that this is a pretty preposterous causal hypothesis; in reality, you would need to use a more sophisticated test of causality (using, e.g., the potential-outcomes framework, mediation analysis, or do-calculus) that controls for the fact that both your X and Y share a potentially causal relationship with a third variable Z.

TL;DR: The statistical analysis method in the paper was testing an overly simplistic, and I would argue incredibly unrealistic, causal hypothesis. While the results are potentially interesting, they also may not be valid. The results are essentially impossible to interpret, to the point of meaninglessness.

There's a reason this isn't published in a respected journal, it's just some university-associated open source one I've never heard of. At least in my field, this paper isn't nearly rigorous enough to be published in any major journal. On any project I've worked on, an analysis like this would be step 1, and would simply be used as a starting point for developing a more sophisticated causal/statistical hypothesis, and then determining the appropriate methods to test that hypothesis. This ... is about on the level of rigor I expect from the first year master's students I supervise.

stormthebeaches
Padawan Learner
Posts: 331
Joined: 2009-10-24 01:13pm

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by stormthebeaches » 2019-07-04 11:30am

But lets go on, because your example betrays your blatant hypocrisy. Lets say Marvel hires hmm, Richard Gere. Someone who China hates. China is under no obligation to allow that Marvel film in. That's freedom to choose what films to allow in their country. Its not fucking censorship.
That is textbook censorship. China would have the right to do it (as would any country) but it is still censorship.

Also, you completely ignored his more serious example of Tiananmen Square.
Oh please. The Left thinks interference in their country is bad, but totally ok for the US to do it to another country. Rachael Maddow feels sorry for John "warmonger" Bolton, because Trump suddenly flip flopped and doesn't want to put in the effort to overthrow Venezeuala's government anymore. That's hypocrisy right there.
Rachel Maddow is one person. Should I bring up The Young Turks, or Jacobin magazine as leftwing media that is against foreign intervention? Should I bring up the Justice Democrats opposition to intervention in Venezuela. Or should I bring up all the anti-war protests by "the left" about Vietnam and Iraq? Maybe you should stop acting like "the left" is a collective entity. Or maybe you shouldn't attempt to derail the thread in some strange attempt to prove "the left" are hypocrites. If you are trying to argue that Chinese censorship is justified maybe you should start another thread instead of trying to derail this one.
Last edited by stormthebeaches on 2019-07-04 11:35am, edited 1 time in total.

stormthebeaches
Padawan Learner
Posts: 331
Joined: 2009-10-24 01:13pm

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by stormthebeaches » 2019-07-04 11:32am

Ziggy Stardust wrote:
2019-07-04 11:22am
I'm going to ignore this petty squabbling to get back to the OP a little.

Now, unfortunately, before I even say anything, I feel it necessary to put my disclaimer at the top. I am NOT trying to be an apologist for Russian interference. Since I know people on this forum have a difficult time responding to arguments without trying to appeal to the political motivations of the poster, I feel I have to point it out. I am ONLY commenting on the study mentioned in the OP from a methodological/analytical viewpoint, to try and better understand the inference it gives us. I have 0 overt political motivations for doing this, with the obvious caveat that my perspective is inherently biased by my own left-wing beliefs. I will also point out, for those of you not aware, I am a professional statistician, so I do have the qualifications to comment on these methods.

(This is quite long, there is a TL;DR version at the end).

Anyway, the elevator pitch for the study is essentially, "Trump's gains in popularity during the 2016 campaign correlated closely with high levels of social media activity by the Russian trolls." More specifically, "The study found that every 25,000 re-tweets by accounts connected to the IRA predicted a 1 percent increase in opinion polls for Trump." (Both these quotes from the article in the OP).

If you look at the original paper for the study, we can dig into the results a little more. For example, we can see the results of one of their primary statistical analyses in Figure 3. This test shows the results of "Granger causality tests" from a "Vector Auto-Regression (VAR)". Now, a VAR can be thought of rather simply as a multivariate time series. Though the devil is in the details, conceptually a VAR isn't much different than a simple linear regression of the kind you learn in algebra (remember y=ax+b?). The "Granger causality test" has a rather grandiose name, but it's just a fancy name for a hypothesis test of a set of variables in the regression model. For example, if your model is y=ax+bz+c, a "Granger causality test" on z would simply be testing that b=0 (the reality is slightly more complicated because in the actual time series case you are dealing with sets of variables, some of which are "lagged" in time, but the intuition is the same). Another way to put it for that same toy model would be to say, "Which of the following equations does the best job of explaining the variation in Y: is it y=ax+c or y=ax+bz+c?" This tells us the degree to which including bz improves the overall explanatory power of the model. In the case of this paper, "Y" is a set of polling results, and "bz" are the social media activities of the Russian bots one week before the polling results. Note that, despite the name, this is not actually a test for "causality", it is (like any regression) a test of conditional correlations. For example, it is unable to properly distinguish between X causing Y, Z causing both X and Y, or Z is caused by both X and Y (again, like any regression).

With that explanation out of the way, let's look at Figure 3 again. It shows us that the Granger causality test found a significant increase in the model's explanatory power for explaining Trump's poll results when Russian bot activity one week before is included in the model. It also finds that there is NO significant relationship between Russian bot activity and Clinton's poll results. This is already something that should raise your eyebrows. If the Russian bots were targeting Hillary to help Trump, why do we not see a significant decrease in her polling numbers as a result of bot activity? We only see an increase in Trump's polling numbers. While polls are technically not a zero-sum game, you expect some degree of symmetry in any changes in polling numbers. Someone only goes up if someone else is going down. But this shows Clinton is not going down. Given the time-frame of the data they are using (2015-2016), you EXPECT Clinton's numbers to remain pretty steady (since she was the favorite to be the Democratic candidate from the start) and Trump's to rise (since he emerged rather surprisingly out of a crowded field of Republican nominees). Unfortunately, the authors do not state whether they controlled for this or how. From a strictly mathematical standpoint this doesn't pose a problem for the Granger causality test, but from an inferential standpoint it also hints at the possibility that the causal model is more complicated then X causes Y, and almost certainly involves another set of variables Z that may be common causes of both.

In fact, digging through their methods, it doesn't appear that the authors controlled for ANY covariates in their model. As the authors themselves note in their abstract, "We find that the release of these tweets parallel significant political events of 2016 and that approximately every 25,000 additional IRA re-tweets predicted a one percent increase in election opinion polls for one candidate." Read that agian, "the release of these tweets parallel significant political events of 2016". They go into some detail in the paper showing that there is a correlation between the occurrence of "significant political events" and increases in Russian troll activity. What they DON'T mention, to their great discredit, is that you would also expect a correlation between the occurrence of significant political events and candidate's polling numbers even in the ABSENCE of Russian trolls. The way their model is currently constructed, they are assuming by default that the ONLY mechanism through which significant political events could impact the candidate's polling numbers is through Russian troll activity on Twitter! It's an unavoidable consequence of their model construction, though they don't explicitly state that. I think on its face that this is a pretty preposterous causal hypothesis; in reality, you would need to use a more sophisticated test of causality (using, e.g., the potential-outcomes framework, mediation analysis, or do-calculus) that controls for the fact that both your X and Y share a potentially causal relationship with a third variable Z.

TL;DR: The statistical analysis method in the paper was testing an overly simplistic, and I would argue incredibly unrealistic, causal hypothesis. While the results are potentially interesting, they also may not be valid. The results are essentially impossible to interpret, to the point of meaninglessness.

There's a reason this isn't published in a respected journal, it's just some university-associated open source one I've never heard of. At least in my field, this paper isn't nearly rigorous enough to be published in any major journal. On any project I've worked on, an analysis like this would be step 1, and would simply be used as a starting point for developing a more sophisticated causal/statistical hypothesis, and then determining the appropriate methods to test that hypothesis. This ... is about on the level of rigor I expect from the first year master's students I supervise.
This is a good post that is actually on topic. Much better than Mr. Friendly guy's ramblings about Chinese's censorship, Venezuela, Rachel Maddow and what have you.

User avatar
mr friendly guy
The Doctor
Posts: 10522
Joined: 2004-12-12 10:55pm
Location: In a 1960s police telephone box somewhere in Australia

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by mr friendly guy » 2019-07-04 11:51am

stormthebeaches wrote:
2019-07-04 11:30am
But lets go on, because your example betrays your blatant hypocrisy. Lets say Marvel hires hmm, Richard Gere. Someone who China hates. China is under no obligation to allow that Marvel film in. That's freedom to choose what films to allow in their country. Its not fucking censorship.
That is textbook censorship. China would have the right to do it (as would any country) but it is still censorship.

Also, you completely ignored his more serious example of Tiananmen Square.
Says the guy who cuts out the rest of my explanation.
Rachel Maddow is one person.
OMG, Rachel Maddow is only one person. Give that man a prize. Maybe I should write "for example" in front of my statement to make it more clear.
Should I bring up The Young Turks, or Jacobin magazine as leftwing media that is against foreign intervention?
Should I bring up that they are the minority in the Left.
Should I bring up the Justice Democrats opposition to intervention in Venezuela.
Should I bring up they aren't the ones holding power in the Left. Get back to me when they do.
Or should I bring up all the anti-war protests by "the left" about Vietnam and Iraq?
Lets play along. How many of these say no interference whatsoever, as opposed to interference via military means.
Maybe you should stop acting like "the left" is a collective entity.
I don't. Troll harder dumbass. No more than when I say <insert country here> I think every single person of that's country's government moves in lock, step and barrel.
Or maybe you shouldn't attempt to derail the thread in some strange attempt to prove "the left" are hypocrites.
Or maybe you shouldn't try to play amatuer mod. If you think its a derail, ask a mod to split the thread. But even you know that topics which get complicated and have numerous side issues arising which don't get counted as a derail.
This is a good post that is actually on topic. Much better than Mr. Friendly guy's ramblings about Chinese's censorship, Venezuela, Rachel Maddow and what have you.
You do realise Ralin and TRR brought China into the discussion right?
Hey using that logic, you're the idiot who brings up the Young Turks and Justice Democrats and antiwar protesters about Iraq. :lol:
Never apologise for being a geek, because they won't apologise to you for being an arsehole. John Barrowman - 22 June 2014 Perth Supernova.

Countries I have been to.
Australia, Canada, China, Denmark, Ecuador, Finland, Germany, Malaysia, Netherlands, Norway, Singapore, Sweden, USA.
Always on the lookout for more nice places to visit.

stormthebeaches
Padawan Learner
Posts: 331
Joined: 2009-10-24 01:13pm

Re: New study confirms that Russian propaganda may have affected the outcome of the 2016 election.

Post by stormthebeaches » 2019-07-04 01:50pm

Says the guy who cuts out the rest of my explanation.
I was only focusing on what I took issue with. Notice how you use that to dodge Tiananmen square. Or my point about that still being censorship.
OMG, Rachel Maddow is only one person. Give that man a prize. Maybe I should write "for example" in front of my statement to make it more clear.
You can't say "the left" believes in something because of the words of one person, which is why I brought in two different left wing news organisations as counter examples.
Should I bring up that they are the minority in the Left.
Rachel Maddow is also a minority in "the Left".
Should I bring up they aren't the ones holding power in the Left. Get back to me when they do.
Two members of the Justice Democrats (Ilhan Omar and AOC) are members of Congress. Does holding power in Congress count as holding power in "the Left". What does it mean to hold power in "the Left" anyway?
Lets play along. How many of these say no interference whatsoever, as opposed to interference via military means.
It's impossible for there to be no interference whatsoever. Large countries can interfere in other countries affairs simply by existing. Trading with another country is technically interfering in their affairs. Even if a country turns to Autarky there will be blackmarkets. And that's just economics. Cultural and political movements in one country can spread to other countries even if that was never the intent.
I don't. Troll harder dumbass. No more than when I say <insert country here> I think every single person of that's country's government moves in lock, step and barrel.
Unless if a nation is in the middle of a civil war it will have a unified government that speaks on its behalf. There is no unified central command ofr "the left". And claiming that "the left" is pro intervention is quite a stretch. Do you have any evidence that interventionist members of "the left" outnumber isolationist members of "the left"? And are we just referring to "the left" in America or "the left" internationally. What do you mean when you say "the left"?
Or maybe you shouldn't try to play amatuer mod. If you think its a derail, ask a mod to split the thread. But even you know that topics which get complicated and have numerous side issues arising which don't get counted as a derail.
My primary purpose is to call you out for attempted derailment, not engage in a debate on that topic. We've gone from talking about Russian interference to talking about "the left", foreign censorship and the alleged hypocrisy of certain board members, thread derailed.
You do realise Ralin and TRR brought China into the discussion right?
Hey using that logic, you're the idiot who brings up the Young Turks and Justice Democrats and antiwar protesters about Iraq.
You are the one who first brought up the issue of Government's censoring Facebook. People assumed you meant China because you have frequently defended China in the past. I brought up the Young Turks, Just Democrats and anti-war protestors as counter examples to you bringing up Rachel Maddow as an example of "the left" being in favour of interventionism.

Post Reply