Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
×
Republicans Twitter Politics

Twitter's Algorithm Favors the Political Right, Study Finds (theconversation.com) 270

An anonymous reader quotes a report from The Conversation: Twitter has on various occasions been accused of political bias, with politicians or commentators alleging Twitter's algorithm amplifies their opponents' voices, or silences their own. In this climate, Twitter commissioned a study to understand whether their algorithm may be biased towards a certain political ideology. While Twitter publicized the findings of the research in 2021, the study has now been published in the peer-reviewed journal PNAS.

The study looked at a sample of 4% of all Twitter users who had been exposed to the algorithm (46,470,596 unique users). It also included a control group of 11,617,373 users who had never received any automatically recommended tweets in their feeds. This wasn't a manual study, whereby, say, the researchers recruited volunteers and asked them questions about their experiences. It wouldn't have been possible to study such a large number of users that way. Instead, a computer model allowed the researchers to generate their findings. [...] The researchers found that in six out of the seven countries (Germany was the exception), the algorithm significantly favored the amplification of tweets from politically right-leaning sources. Overall, the amplification trend wasn't significant among individual politicians from specific parties, but was when they were taken together as a group. The starkest contrasts were seen in Canada (the Liberals' tweets were amplified 43%, versus those of the Conservatives at 167%) and the UK (Labour's tweets were amplified 112%, while the Conservatives' were amplified at 176%).

In acknowledgement of the fact that tweets from elected officials represent only a small portion of political content on Twitter, the researchers also looked at whether the algorithm disproportionately amplifies news content from any particular point on the ideological spectrum. To this end, they measured the algorithmic amplification of 6.2 million political news articles shared in the US. To determine the political leaning of the news source, they used two independently curated media bias-rating datasets. Similar to the results in the first part of the study, the authors found that content from right-wing media outlets is amplified more than that from outlets at other points on the ideological spectrum. This part of the study also found far-left-leaning and far-right-leaning outlets were not significantly amplified compared with politically moderate outlets.
The authors of the study point out that the algorithms "might be influenced by the way different political groups operate," notes The Conversation. "So for example, some political groups might be deploying better tactics and strategies to amplify their content on Twitter."
This discussion has been archived. No new comments can be posted.

Twitter's Algorithm Favors the Political Right, Study Finds

Comments Filter:
  • Well ... (Score:5, Funny)

    by Anonymous Coward on Monday January 31, 2022 @06:22PM (#62225145)
    So much for #rigthwing #victim #censorship #uuunfaair #whining ...
    • First, twitter being biased is an entirely different claim than twitter's automated system being biased. Most people who complain about bias are talking about what twitter does to step in and manually alter things. Second, "conservative" in the US, UK, and Canada are often different (as is "liberal") and aren't at all the same thing as "right-wing" or "left-wing." Third, at the end it pretty much states that they don't actually know if their algorithms are biased or not, and that certain groups are better
      • The U.S.A. seems to be the only country where “liberal” is a social term rather than a fiscal one and pertains to the identity politics rights of certain groups.

        In most of the world, “liberal” is synonymous with “capitalism”, id est less rules or companies. — The U.S.A. has a tendency to develop political lexicon separate from the entire world, indicative of the isolationist culture that does not receive a lot of input from outside of it./p.

    • Re: Well ... (Score:5, Insightful)

      by klipclop ( 6724090 ) on Monday January 31, 2022 @07:30PM (#62225401)
      I'm pretty sure their algorithms favor conspiracy theories, stupid narratives and anything that will annoy the average person who has an iota of critical thought. They just want to keep you on the platform as long as possible. It just so happens to be the far right (and far left). If it was socially acceptable to show babies being drowned and kept eyeballs and "engagement", twitters algorithm would happily favor that too....
    • Anything that discredits the conspiracy is just part of the conspiracy. Duh!

    • by Ungrounded Lightning ( 62228 ) on Tuesday February 01, 2022 @02:48PM (#62227711) Journal

      So much for #rigthwing #victim #censorship #uuunfaair #whining

      The study had NOTHING to do with the CENSORHIP algorithms.

      It was studying the SILOING algorithms - which select among the (post-censorship remainder of the) universe of posting presented to the user, to see if liberal versus conservative viewers got more amplification.

      It found a substantial focusing (about 2x for US and Japan to about 3x for Germany. Given the way they lumped parties into "left" and "right", virtually all the countries showed slightly more increase in political shift due to focus for users identified as right than as left partis. Outliers were Germany (TINY amount stronger for left than right) and Canada (big shift for right, quite small for left.)

      These results are completely consistent with this explanation: The universe of (post censorship) postings has more left-wing than right-wing posts, so an unbiased show-em-what-they-line algorithm produces more adjustment for right-wing than left-wing readers.

      That, in turn, would be consistent with either (or both) of TWO explanations:
        1) Left-wingers (and astroturf operations) post more than right-wingers (and astroturf operations)
        2) Censorship removes more right-wing than left-wing posts.

      So to the extent it says anything about censorship it suggests the right wing is more censored.

  • Well, now (Score:4, Insightful)

    by 93 Escort Wagon ( 326346 ) on Monday January 31, 2022 @06:26PM (#62225165)

    This discussion thread should be entertaining...

  • by SuperKendall ( 25149 ) on Monday January 31, 2022 @06:27PM (#62225167)

    The researchers found that in six out of the seven countries (Germany was the exception), the algorithm significantly favored the amplification of tweets from politically right-leaning sources.

    That's a lot of words to express a simple truth:

    The Left Can't Meme.

    • by fahrbot-bot ( 874524 ) on Monday January 31, 2022 @06:37PM (#62225217)

      The researchers found that in six out of the seven countries (Germany was the exception), the algorithm significantly favored the amplification of tweets from politically right-leaning sources.

      That's a lot of words to express a simple truth:

      The Left Can't Meme.

      More likely, that those on The Right are less critical and/or independent thinkers.

      • That's the entire point. Memes aren't high level thinking. They are usually silly or dumb but they make you chuckle and they stick for whatever reason. A lot of memes are quite slapstick.

        Take for instance, Let's Go Brandon!

        It originated from a nascar, or some sporting event, where the crowd was chanting Fuck Joe Biden. Well the announcer was on broadcast television and for some reason they acknowledged the crowd and the guy says, I think they are saying Let's go Brandon!! Who the fuck is Brandon and are you

      • More likely, that those on The Right are less critical and/or independent thinkers.

        Something like this goes through my mind every time I see my right-wing-leaning family/friends all high-fiving each other because their kids are growing up "right" as they are learning how to sing Springsteen's Born in the USA. I cannot fathom how many times they've heard the song, yet they still do not seem to even begin to grasp what it's about. Might as well be tossing around Lopez's Booty.

      • Re: (Score:2, Troll)

        by JBeretta ( 7487512 )

        More likely, that those on The Right are less critical and/or independent thinkers.

        You have got to be kidding me. The left is all about Group Think and Group Identity.

        • by GlennC ( 96879 )

          The left is all about Group Think and Group Identity.

          And how is this different from the "Right"?

          Face it, both sides are "Yay Us, F Them!"

    • by _xeno_ ( 155264 )

      It's also funny because I've seen Twitter try and manually override the algorithm by attaching an "event" to some hashtag that the right got trending. Something like #BareShelvesBiden suddenly becomes a "Twitter event" explaining that, no, the shelves aren't really bare, ignore your eyes, the supply chain is actually just fine, nothing to see here.

      It's not impossible to see the actual tweets when they attach an "event" to them, but it becomes much more difficult, because the "event" replaces the "trend."

    • and is more willing to use dirty tactics like paying people to up/down vote everything they can in line with the rights viewpoint.
  • by Huitzil ( 7782388 ) on Monday January 31, 2022 @06:27PM (#62225169)
    Our definition of âoeright wingâ today has evolved. Itâ(TM)s not Reaganism that we are talking about here - itâ(TM)s bow tie wearing firebrand talking heads that say Outrageous Things that end up making any collaborative filtering algo bubble posts to drive engagement. People like to read Outrageous Things. This is why Murdoch is so successful.
  • Help (Score:4, Funny)

    by ArchieBunker ( 132337 ) on Monday January 31, 2022 @06:27PM (#62225173)

    We're being oppressed!

  • Echo effect? (Score:5, Interesting)

    by Tablizer ( 95088 ) on Monday January 31, 2022 @06:28PM (#62225179) Journal

    This could also mean the right is more likely to accept and forward BS. Many believe they reject rational thought out of culture and upbringing. [blogspot.com]

    • Re: (Score:3, Informative)

      by AmiMoJo ( 196126 )

      Similar findings were made about Facebook, and the main reason was that conservatives seem to be less likely to check facts or question things before liking and sharing.

      • I've never seen any solid, independent research that shows either "party" being less prone to mindless bullshit than the other. What I have seen, is that people who do NOT align with a political party are far less likely to spread bullshit... regardless of whether their political stances are left, right, or center.
        • I ran across a study that examined how different political groups spread "low-credibility content". Groups belonging on the center to the far left are usually better at questioning if something is true or not, while groups belonging on the right were more prone to accept "low-credibility content". The groups on the right were also more prone to attacking their own if someone from their group questioned the collective wisdom of what was true or not.

          You don't have to look further than to the US Senate to see

          • I ran across a study that showed people believe lies they want to be true.
            • I ran across a study that showed people believe lies they want to be true.

              Oh, do provide a link to that study so we can compare it to the study I referred to, since it says much the same. You can read it yourself, it's called Network Propaganda: Manipulation, Disinformation, and Radicalization in American Politics, link https://oxford.universitypress... [university...arship.com]

              Here's a little excerpt:

              The consistent pattern that emerges from our data is that, both during the highly divisive election campaign and even more so during the first year of the Trump presidency, there is no left-right division, b

        • I've never seen any solid, independent research that shows either "party" being less prone to mindless bullshit than the other.

          * I'm not sure how solid, but there is: Why Liberals Are More Intelligent Than Conservatives https://www.psychologytoday.co... [psychologytoday.com]
          * Not meaning who's more clever than the other, but their brains are different: "greater liberalism was associated with increased gray matter volume in the anterior cingulate cortex, whereas greater conservatism was associated with increased volume of the right amygdala. [...] Our data do not determine whether these regions play a causal role in the formation of political attitudes

    • Re: (Score:3, Interesting)

      by argStyopa ( 232550 )

      I *regularly* get tweets from a friend "can you fucking believe this?" with some stupid shit Trump said.

      So he's forwarding some right wing meme, does that count as a right or left-wing 'tweet'?

    • by taustin ( 171655 )

      Or it could mean that the right is so accustomed to censorship that they feel the need to fight against it by engaging the Streisand Effect [wikipedia.org].

      Speculating is fun, and you can do it without actually knowing anything.

    • by Xenographic ( 557057 ) on Monday January 31, 2022 @08:39PM (#62225607) Journal

      > This could also mean the right is more likely to accept and forward BS. Many believe they reject rational thought out of culture and upbringing. [blogspot.com]

      If only there was some way to figure out what the study was about [pnas.org], perhaps by clicking on links in the fine article, you could show off that rational thought that's implicitly a part of your culture and upbringing. Then you might realize that the only mention of replies is in how they excluded politician's retweets because attribution becomes ambiguous and they treat the algorithm as a black box, so there's nothing in there showing such effects at all. Here's what they did find:

      We present two sets of findings. First, we studied tweets by elected legislators from major political parties in seven countries. Our results reveal a remarkably consistent trend: In six out of seven countries studied, the mainstream political right enjoys higher algorithmic amplification than the mainstream political left. Consistent with this overall trend, our second set of findings studying the US media landscape revealed that algorithmic amplification favors right-leaning news sources. We further looked at whether algorithms amplify far-left and far-right political groups more than moderate ones; contrary to prevailing public belief, we did not find evidence to support this hypothesis.

      This amplification they're measuring was basically trying to see if Twitter's algorithm thought that right-wing or left-wing politicians' tweets were more relevant to the users on average. For whatever reason, the algorithm thought that right wing politicians' tweets were more relevant to people. They do not (and say that they cannot) identify any causal factors in this and despite apparently getting this data from Twitter, they haven't actually analyzed how the algorithm works as anything but a black box. If there was any pre-registration for this study, I didn't find any, nor do we know what other questions they analyzed and didn't find significant results for. Given that this is a rather empty clickbait finding, that's a bit sus.

      Then they go on to mention this:

      When studying amplification at the level of individual politicians (Fig. 1C), we find that amplification varies substantially within each political party: While tweets from some individual politicians are amplified up to 400%, for others, amplification is below 0%, meaning they reach fewer users on ranked timelines than they do on chronological ones. We repeated the comparison between major left-wing and right-wing parties, comparing the distribution of individual amplification values between parties. When studied at the individual level, a permutation test detected no statistically significant association between an individual’s party affiliation and their amplification.

      This makes it seem like the content, not the party, is what really matters. And the amplification isn't necessarily from people who agree with them, it may just as well be from outrage clicks. Alas, they don't really go into any of that and the study mentions in their limitations that it basically sees an effect but has no idea why, so it's honestly pretty worthless.

      But hey, some of us had the culture and upbringing to actually read the study before commenting instead of going for snarky comments first. Sometimes it's best to save those for after :)

  • I assumed the Twitter userbase was just more left. At least, that's how it seems when I've looked at replies to politician tweets.

  • "Twitter's Algorithm Favors the Political Right, Study Finds " the platform favors those getting censored, silenced and booted off.
  • Now who should we believe?

  • by PPH ( 736903 ) on Monday January 31, 2022 @06:45PM (#62225247)

    For many years (decades), most of the talk radio has had a rightward lean to it. Left wingers tried to launch a "progressive" talk radio station in Seattle. But it died due to lack of interest. In Seattle.

    There just doesn't appear to be much of a business case for left wing broadcasts and other material. And since Twitter is in business to attract eyeballs, it will drift right eventually. That's where the public is.

    • For many years (decades), most of the talk radio has had a rightward lean to it. Left wingers tried to launch a "progressive" talk radio station in Seattle. But it died due to lack of interest. In Seattle.

      There just doesn't appear to be much of a business case for left wing broadcasts and other material. And since Twitter is in business to attract eyeballs, it will drift right eventually. That's where the public is.

      ...everyone else wants to actually be right. That's why liberal media fails. Liberals want to actually be correct. They want their worldview to line up with the facts. When the facts contradict their worldview, it's very distressing. Also, most of us the Right Wingers call "liberals" or "the left" are just centrists or people who find modern Right Wing ideology repellent. For me, for example, put me in a room with real far-left liberals, I'll tell them they're fucking idiots and it won't be pleasant.

    • Re:Talk Radio (Score:5, Insightful)

      by narcc ( 412956 ) on Monday January 31, 2022 @09:08PM (#62225667) Journal

      It's not the the public is right-wing. Far from it. The public overwhelmingly support progressive policies, including Republicans. It's just that right-wing radio is so much more exciting.

      Left-leaning radio is boring. It's little more than just the news with some predictable commentary. It's really hard to manufacture outrage when your listeners are informed and care about the truth.

      Right-wing radio, on the other hand, is crazy exciting. Secret government plots, shadow organizations, clones and body doubles, ancient conspiracies and secret knowledge, nefarious plots and secret agents ... and that's just the mainstream stuff.

      Remember Weekly World News? That's the tabloid with aliens and Bat Boy, not celebrity gossip. That was a lot of fun to read. Right wing talk radio is a lot like that, but with the fun replaced with fear and hatred.

      Just imagine if the pandemic was a movie.

      The left wing version goes like this: "Oh no! A plague! Here are a few basic things we can do to reduce our risk. Oh, look! Our scientists have made a vaccine! Let's all take it."

      The right wing version is a bit more dramatic: "Oh no! The government is killing people to make it look like there's a plague! They're forcing us to wear masks to condition us into accepting total control over every aspect of our lives! The vaccine contains microchips that will track you and modify your behavior, and also kill everyone who takes it within 3 years as a means to drastically reduce the population! Also, the plague is actually real but engineered by our top-scientists secretly colluding with China for some reason. We know the cure is to take a mix of various drugs from the feed store, Viagra, and hormone pills along with a selection of household cleaners and our own urine, but they're using ventilators to kill us in the hospital to keep us from sharing the secret we've already posted all over the internet."

      Right-wing radio is wild. Not only does it make our boring old world much more exciting, you get to pretend that you're smarter / more knowledgeable than everyone you know about, including our nations top scientists! It's no wonder it bring in listeners.

      • by splutty ( 43475 )

        So the History Channel is right wing talk TV?

  • Interesting findings. I wonder what they actually mean. Lets try to break this down.

    Here are some general observations, so before you all totally flame me on ism's or grouping people, just remember, I don't pretend to be a robot, just a human with a sense of humor.

    "Liberals" try to claim that they aren't mean. They only rely on "science". They tend to say that they don't do edgy (race based, skill based, personal attribute) jokes. Generally, when I read this stuff, it's all fairly boring, accurate o

  • by Mononymous ( 6156676 ) on Monday January 31, 2022 @07:08PM (#62225329)

    "Receiving automatically recommended tweets in your feed"? What a horrible idea!

    I'm sure glad that I use Fritter to read the tweets of people I'm interested in.

  • But, hey. We have a paper now. You can google the answer you wish to get.
  • But not the CEO...

    • The ex ceo favored free speech. The main target of censorship at the moment is right wing speech. Liberal attitudes are genereally pro censorship.
      Therefore rightwing speech will flock to twitter as a remaining place which supports free speech . That makes it even more of a target.
      The new CEO will surely change all that.

      The next step is that people gradually find out that censorship is far more reaching than right wing. Duh.

  • People should spend their time doing something useful, like cleaning the pizza boxes up off the floor.
  • But we've known this for ages. The right wing consistently have the top 10 politics posts on Twitter. And it's not that we don't understand why. They have a ton of money and they spend it to manipulate all forms of media. Also it's much easier to make something on the right go viral because the right wing will accept absolute nonsense while the left wing will debunk it pretty quickly.

    This isn't me insulting the left or right. This comes from people who sell ads for a living and use fake news to direct p
  • Amplification comes from Echo Chambers. You fill in the blanksâ¦
  • When Amazon's stops recommending books I've already bought from Amazon, and YouTube stops recommending sports videos when I have never, not even once, watched a single sports video.

    I doubt that the people behind this are actually smart enough to fake a study as part of tighter censorship of the right, but when they act like that's what they're up to, a lot of people will assume that is, in fact, what they're up to.

    • When Amazon's stops recommending books I've already bought from Amazon ...

      You have to train your Amazon AI. I have got mine to recommend interesting and possibly subversive books about economics, anthropology, etc. I think I got the AI a bit confused by buying Adam Smith's "The Wealth of Nations" and "The Communist Manifesto" by Marx and Engels in one order. After buying a collection of essays by John Stewart Mill, maybe the AI got a clue that I might be a bit old style liberal and reformist. "Capitalism and Freedom" by Milton Friedman was a frequent recommendation, but I haven't

  • How many Twitter accounts have been cancelled or suspended, and why?

  • We define the reach of a set T of tweets in a set U of Twitter users as the total number of users from U who encountered a tweet from the set T.

    ...where T are the tweets of elected politicians of a liberal or conservative bent, and U are the total number of users across either a control group (no algorithm boosted tweets shown) or the experimental group (who saw the algorithm boosted tweets when Twitter revamped their feed). The study, simply put, asks if the algorithm is more likely to show you conservative or liberal tweets from politicians?

    Now consider two facts:
    1. Twitter's user base is significantly more liberal than average. (Source: Pew Re

  • And what is the political bias of the authors of the study?
  • I recently read a rather interesting bit of archived journalism (Foreign News, 1947), about the development of soviet communism under Stalin. The telling statement was that if you believe that the rest of the world is against you, and you act on that belief, then pretty soon, the rest of the world will prove you right. Stalin was convinced that there were reactionary forces hostile to the communist revolution, which had to be rooted out within Russia, and opposed where they originated from the decadent capi

  • by splutty ( 43475 ) on Tuesday February 01, 2022 @05:54AM (#62226383)

    News at 11.

  • Deny, obfuscate, lie. Pretty standard behavior from our wise masters.
  • The research seems to say that negativity increases engagement and the right's entire platform is, literally, negativity. I'm very much not on the political left, I'm just being realistic.

    The republican party no longer has a platform based on any principles or ideology. Ideals like individual freedom and personal responsibility are completely absent. In the case of fiscal responsibility, they have become demonstrably worse than the democrats.

    What the republican party has become is "not the democrats". W

"I got everybody to pay up front...then I blew up their planet." "Now why didn't I think of that?" -- Post Bros. Comics

Working...