Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Communications The Internet Network Networking News Entertainment Technology

The Guardian Publishes Comment Abuse Stats, Invites Debate On Moderation (theguardian.com) 303

AmiMoJo writes: British newspaper The Guardian has published some stats on its popular comment sections attached to each story. So far the Guardian's site has received 70 million comments, of which around 2% were removed for violation of community standards. Articles written by women tended to get the most blocked comments, especially if they were in male-writer dominated sections like sports and technology, while fashion was one of the few areas where men got more abuse. Further down the article the reader is invited to moderate some sample comments and see how their actions compared to those of the paper's staff. You can leave suggestions for improvement here.
This discussion has been archived. No new comments can be posted.

The Guardian Publishes Comment Abuse Stats, Invites Debate On Moderation

Comments Filter:
  • should concentrate on their own credibility first, then worry bout their posters.

    • by Anonymous Coward

      ... around 2% were removed for violation of community standards ...

      I have given up on commenting there because I know how they rate their 'community standards'

      It was an article on the flood of millions of migrants into Europe and I was asking if it is wise to allow so many of those who have no intention of integrate into the European culture ... and my comment was nuked

      There was no cussing

      There was no degrading of any specific race

      There was not even any mention of religion

      I was only commenting on the wisdom of allowing so many of those who had shown to not interested of in

      • by Anonymous Coward

        Same experience on Scientific American. No violation of their own stated policies. Not even mentioned global warming.
        Comment deleted.
        I complained and their reply was that I lost my "privilege" to post.
        I cancelled the renewal of my subscription and unsubscribed from any of their mailing lists.
        Absolutely disgraceful.

      • by guestapoo ( 4136621 ) on Wednesday April 13, 2016 @02:38AM (#51897583)
        ...and it's annoying that, if you be banned your comments' history will be gone also, they are vapored without any warning. I were one of the 'victims'.
        And, after that I found out many readers have the same problems with this 'moderation system', no matter they are left or right, and they joke that 'CiF (Comment is Free, is what The Guardian call their comment section), but some comments are freer.

        And, their 'community standard' is very obscure, while some personal attacks are still there, some normal debates were deleted. When I demanded they show how I violated the rules (yes, with "please"), they deleted my comment also.

        It's not that I'm whining, I'll be fine that I must respect the 'rules', I will be OK if, 'you violated 'Section 1, clause 3', so your comment was deleted', but in The Guardian there is no rule. It's likely the moderators delete comments they don't like.
      • Re: (Score:2, Insightful)

        by dave420 ( 699308 )

        Your comment was nuked because you made a baseless claim that they don't want to integrate. That is degrading. If you can't see why, you should get a refund on your education, as it failed you massively. If you could post the verbatim full text of your comment we could get to the bottom of this. As it is it sounds like your lack of critical thinking and desire to leap to the inaccurate led to the nuking, which is fair enough. I guess the Guardian doesn't want their comments section filled with idiots m

        • My wife (a doctor) sees patients (both male and female) who have lived in this country for 15 - 20 years, and they still need a translator because they haven't bothered to learn the local language. Some people simply don't want to integrate.

          • Some != all.

          • by GNious ( 953874 ) on Wednesday April 13, 2016 @05:58AM (#51898155)

            (I'm in Belgium)
            When governments and societies take foreigners and group them together geographically, you end up with small enclaves where there's ca no reason or need to learn the national language.
            I can show you places in here in Belgium, where Arabic will help you more than French or Flemish, there are sections in Frankfurt where you might as well speak Turkish, large area in northern Mexico where half the population only speaks german and sections of Spain where no-one speaks anything but UK English.

            Yes, some people have no interest in integrating, but when there's no way to even use the national language, at lot of foreigners end up giving up trying to learn it.

        • by Raenex ( 947668 ) on Wednesday April 13, 2016 @04:08AM (#51897861)

          Get real. Comments that go against the agenda are nuked for any or no reason beyond that they go against the agenda. The standard you're applying is only applied to one side of a position. One thing you can say about Slashdot, your comment may be moderated to the basement by groupthink, but at least it's there for people looking.

    • Re: (Score:2, Informative)

      by Anonymous Coward

      I went through their list of would you allow or block thing, and wow. Most of the comments I looked at and said, "well, I disagree with your opinion, but you're free to it and it's not overly trolly so allow", and they blocked almost all of those. It sort of felt like censorship of anything that didn't fit their view point. If that's how they run their news site, I actually kind of get your statement on credibility.

    • by Mashiki ( 184564 ) <mashiki&gmail,com> on Tuesday April 12, 2016 @10:08PM (#51896751) Homepage

      Well they've got to go down this path bu choice, they're only 5 years out from complete bankruptcy that includes the money from the trust. They need to try showing people that they know how to remain solvant and can get their head on and fix the problems...and see...those problems? We want your opinion!

      Personally, I think them having Jessica Vallenti was a great idea, it was a really good comedy section. I mean look at these examples: One week she gets to tell the idiots that wolf whistling is sexist, a few weeks later, she says it's sexist not to wolf whistle at someone. Then she starts going off about how air conditioning is sexist, but it's really not sexist all in the span of 3 weeks.

      • by AmiMoJo ( 196126 )

        I think the Guardian is a bit beyond your reading comprehension level, if that's what you got from those articles.

        I'm not a fan or regular reader of The Guardian, but I can see and respect what they are trying to do. Newspapers are dying and they are trying hard to go with the change, having picked a particular niche that does seem to be under-served. That is, high quality journalism and a high quality debate with fairly strict moderation, somewhat like a live debate would have.

        Of course you can argue if th

  • Opportunity missed (Score:5, Insightful)

    by Gravis Zero ( 934156 ) on Tuesday April 12, 2016 @10:02PM (#51896739)

    What they failed to do was publish articles written by one gender/race under the alias of the opposing gender/race, serving it to part of the internet and seeing if there is a difference in the number of negative responses. They also didn't track the geographic region of the originating blocked posts. No conclusions can be taken from these numbers besides, "some people on the internet are assholes" which we already knew for a certainty thanks to the youtube comment sections.

    • by swb ( 14022 )

      The gender/race alias thing would be genius.

      It'd be so easy to do online you almost wonder why they haven't done it already, unless they're worried about what theories it might not validate.

      • I see the dismissal and accusations of cherry picking are here already. The point readily highlighted by their data is that people seek out and shit on people based on race and gender. It's unlikely the A/B testing the GP cited has been done, as it'd be difficult and is somewhat outside the scope of a newspaper.

        • I haven't dismissed their findings, I'm simply saying that any conclusion drawn from this data is premature, aside from the conclusion that some people on the internet are assholes.

          • So far as it goes, it seems like a reasonable conclusion to draw. They have found a pretty strong statistical correlation between writers who are female or non-white. If you read some of the analysis, there is some good points that there may be some selection bias going on, particularly as some moderation doesn't appear to follow the Guardian's community standards, but even with that in mind, there's a strong enough signal there to suggest that women and non-whites who contribute to the Guardian are more li

            • That may be the case, but you can't then come out and say it's because they're women and ethnic minorities, which is what the article and the berk who submitted it are trying to imply.

              • by Sique ( 173459 )
                No, all the article does is to point out the strong correlation. The "because" is something you read into it. I wonder why you might feel so inclined to conclude that the correlation pointed out in the article is actually a causation.
                • I didn't say that it said there was causation. I said it implied it. You can imply things by not saying anything.

                  The writer of the article probably knows that most Guardian readers graduated in underwater basket dancing and will form the required conclusion anyway.

                  • by Sique ( 173459 )
                    This is a handwaving argumentation. "Why would someone point out a correlation if he didn't want to hint at a causation?"

                    That's your opinion about the article. That's not what the article states. All the article does is saying that they did some data mining in their 70 mio comment database and found some strong correlations, which they list in the article. It's solely you who concluded that the article somehow constructs a causality.

                  • by Maritz ( 1829006 )

                    The writer of the article probably knows that most Guardian readers graduated in underwater basket dancing and will form the required conclusion anyway.

                    Laying on the irony thick there. Bias is only bad when it's the wrong kind of bias, eh? lol.

        • by swb ( 14022 ) on Tuesday April 12, 2016 @10:35PM (#51896869)

          Wait, you mean "going undercover" is a new thing that journalists don't know how to do? They never ghostwrite anything or use pseudonyms or otherwise mask their identity? And I'm sure it would be *so hard* to mask an identity on the Internet, I mean, nobody gets away with that. On the Internet, everybody knows you're a dog.

          While it doesn't entirely surprise me that columnists identifiable as specific genders or races might attract more negative comments, it'd be illuminating to have data on how often they write any kind of advocacy journalism or, and if, and to what extent they make their race or gender part of their subject matter.

          My guess is that high visibility blacks and women are highly correlated with either controversial opinions and/or controversial subjects. And that hostile commentary is highly correlated with controversial opinions or subjects.

          It seems less plausible that blacks or women who write in identity free tone about uncontroversial topics will attract identity-focused hostile comments.

          • by AmiMoJo ( 196126 )

            On the Internet, everybody knows you're a dog.

            While that is an option, most journalists prefer to have their name attached to their stories because it's how they build up their careers

            It seems less plausible that blacks or women who write in identity free tone about uncontroversial topics will attract identity-focused hostile comments.

            That's kind of the point. People shouldn't have to hide their race or gender to avoid the abuse those attributes seem to attract.

    • by AmiMoJo ( 196126 )

      I wish they had gone further too, but this does provide solid evidence for a couple of important points.

      1. There is a real gender bias. It's mostly against women, but in a small number of categories it's against men.

      2. This bias may be partially accounted for by the feeling that a predominantly male/female space where the commentator feels safe and comfortable is being "invaded" by the other sex.

      It's also interesting that there seem to be certain trigger words that some people developed a Pavlovian response

      • "It's also interesting that there seem to be certain trigger words that some people developed a Pavlovian response to. I think that deserves a lot more research."

        Gotcha. I find it often. You just need to write certain keywords that quickly someone writes a crazed response. As example in my country if you just type "communism", promptly someone will want to kill you as if we were still in the cold war.
  • violation of community standards

    What's with the "community"? There are no "community" standards — the removed messages were deemed offensive by a handful of moderators. Moderators prone to keeping some posts more equal than others [dailymail.co.uk] and susceptible to manipulation by evil regimes [theguardian.com].

    Calling them "community" is redefining terms [wikipedia.org]...

    • by starless ( 60879 ) on Tuesday April 12, 2016 @11:50PM (#51897191)

      For what it's worth, my comments to the Guardian are these days are now in the category of: "Your comments are currently being pre-moderated".
      This is because I disagreed (in a fairly reasonable way - certainly not aggressive and not obscene) with the remarks of a British reporter, based in New York, who was making remarks about the situation in my home town of Baltimore. In that case it appeared that the reporter himself may well have been the person removing the negative comments. The fraction of comments removed from all posters for that article was a huge fraction of all those posted. However the Brit reporter (who may not even have visited B'more ) was apparently very sensitive to criticism - even if it wasn't that different from the criticism he was dealing out to some of the locals. One factor may be the "knee-jerk" anti-Americanism of certain segments of British society - who are surprised to receive similar criticisms back themselves.

    • It also happens that the standards applied to comments are stricter than those applied to the articles, themselves.

      Authors can and do get away with expressing outrageous opinions, not just ones that are factually incorrect, but offensive remarks, sexist and on rare occasions statements implying approval of physical abuse. If they gad appeared in the comments, they would have been removed. But moderation is not applied to their authors.

  • by Etcetera ( 14711 ) on Tuesday April 12, 2016 @10:18PM (#51896793) Homepage

    Or at least... a Slashcode-like commenting, moderation, and meta-moderation system.

    For all we complain about it here, and for all the trolling that occurs, the Slashdot moderation system seems to have passed the test of time reasonably enough.

    Perhaps it's a little like that infamous definition of Democracy: It's the worst form of government except for all the others that have been tried.

    • by AmiMoJo ( 196126 ) on Wednesday April 13, 2016 @03:36AM (#51897757) Homepage Journal

      Slashdot's only real weakness is that people, especially groups of like-minded people, can use the -1 mods to try to silence people they disagree with. While I think having down-mods is a good thing generally, once a post gets a single +1 all the -1s should only count for -0.1 each.

      Controversial != troll.

      • Slashdot's only real weakness is that people, especially groups of like-minded people, can use the -1 mods to try to silence people they disagree with.

        Agreed. The system does very well with technology and science articles, as there's typically little argument over the facts themselves. However politics articles are highly prone to groupthink, with libertarian-leaning comments tending to do better than anything else (and heaven help the Trump supporter!).

  • Diversity hires (Score:2, Interesting)

    by Anonymous Coward

    It looks like The Guardian just got a first-hand lesson on the flaws of "diversity". When allow racism and sexism to dictate your hiring policy, you'll inevitably get people who wouldn't have been hired on their merit alone.

    I wonder how much of the "abuse" was of the "you suck" [youtu.be] variety. Maybe The Guardian needs to ask the UN to censor their critics as well.

  • by sethstorm ( 512897 ) on Tuesday April 12, 2016 @11:33PM (#51897107) Homepage

    Unlike The Guardian, Slashdot doesn't answer to political grievance groups (and has only removed one thing for Scientology).

    I'm not sure that this is debate as much as it is a justification.

  • by ScentCone ( 795499 ) on Tuesday April 12, 2016 @11:46PM (#51897177)
    It's not about the gender of the author, it's about the agenda of the author. Regardless, most flamebait/troll posts are aimed at other comments anyway - at least that's apparent in watching this play out in other papers' comment areas. Also, the straight reporting or analysis vs. op-ed will make a huge difference.
    • Re: (Score:2, Insightful)

      by AmiMoJo ( 196126 )

      It's not about the gender of the author, it's about the agenda of the author.

      How do you figure that? It sounds like you are saying female authors troll their readers more.

      • How do you figure that? It sounds like you are saying female authors troll their readers more.

        Some do in the Guardian. Jessica Valenti, for example, is a notorious troll.

        In TFA she is quoted as saying:

        'Imagine going to work every day and walking through a gauntlet of 100 people saying "You're stupid", "You're terrible", "You suck", "I can't believe you get paid for this". It's a terrible way to go to work'

        Hundreds of people say this, because it's true.

  • by petes_PoV ( 912422 ) on Wednesday April 13, 2016 @12:59AM (#51897377)
    The newspaper has an article that explains a little of how they analysed their data (Postgres, cloud, Perl). However it does appear that they have "valued what they've got", rather than getting what they value.

    So, for example, they cite that 8 or their 10 most "abused" contributors were women (4 white, 4 not). But since the newspaper has a great deal of coverage of "women's" issues including a lot of highly opinionated articles about feminism -- but no corresponding articles, or sections, concerning men's issues their coverage is neither balanced, nor able to show how much abuse their "men's" writers would get, since they don't have any.

    They also counted all "moderated" comments (ones that do not meet their community standards) as being abusive and they assume that the abuse is directed against the author. However, they remove comments that are off-topic and ones that make personal comments about other commentators. So a comment that was removed because it insulted a commentator who was attacking the article (i.e. the insulter was supporting the author), would be counted as abuse against that author. They do not give reasons for removing comments and only have the single classification.

    Finally, The Guardian admits that it does not moderate either consistently: applying different levels of rigor depending on the topic, nor does it moderate all articles to the same extent. It also does not open all it's articles up to comments.

    In all, while their analysis does point to there being online abuse - they reckon they delete 2% of articles, from the 70 million submitted so far, the results are patchy, inconsistent and cherry-picked. It would never pass a peer review and seems to have been published more to push the newspaper's own agenda, rather than as an authoritative work to highlight a problem (they don't say if the level of "abuse" is rising or falling since they started in 1999).

    • Peoples lack of attention to details like you've described seem to be a considerable problem.

      I find commonly the responses from going into detail into the problems with articles like this are "YOU HATE WOMEN" and other things along those lines.

      I've yet to find a way to get people actually interested in the details, rather than spewing forth knee-jerk hatred.

    • The newspaper has an article that explains a little of how they analysed their data (Postgres, cloud, Perl).

      I think I found the problem!

    • by AmiMoJo ( 196126 )

      The Guardian seems to be trying to set a high standard for comments in order to present itself as a place for serious, insightful debate. They obviously haven't seen Slashdot, which generates insight by tolerating pretty much anything within the bounds of the law.

      They also counted all "moderated" comments (ones that do not meet their community standards) as being abusive and they assume that the abuse is directed against the author.

      That's not quite right. They aren't saying those comments are abuse, they are just saying that they were moderated (e.g. for being off-topic). Certain topics and certain genders tends to have more comments moderated, meaning there is either more ab

    • by T.E.D. ( 34228 )

      of highly opinionated articles about feminism -- but no corresponding articles, or sections, concerning men's issues their coverage is neither balanced, nor able to show how much abuse their "men's" writers would get, since they don't have any.

      That would be a total false equivalence. If group A runs most everything and group B does not, then arguing for more rights and privileges for group B is a qualitatively different thing than arguing for more rights for group A. You could only fairly make that comparison if everything was both now and historically equivalent for both groups.

      I don't think this should be a difficult concept. You don't see people get nearly as angry at kids punching adults in public as they do at adults punching kids. The pow

  • by Anonymous Coward

    Those comments that they showcased in the article were in many cases perfectly legitimate debate.

    The comment was:
    “A 12-year-old boy, out at night, waving a BB gun? What sort of parent allows that? What happened is the product of a fucked up
    society/community/culture/upbringing. I'm sorry to say, but often black people are their own worst enemies.”

    You answered allow. We thought differently.
    This was removed for racism (“black people are their own worst enemy”; “fucked up community

  • Articles written by women tended to get the most blocked comments, especially if they were in male-writer dominated sections like sports and technology, while fashion was one of the few areas where men got more abuse.

    How did the articles on sports/technology fashion fare?

  • Why are they censoring comments? That's just comments for chrissake! Comments reflect people's opinion, all people, and ensures the reader to get a fair sample of what people think. So the Guardian may censor that comment, that I would have seen as interesting, and may keep this one that I may feel too politically-correct. It's a website, keep all comments! Add a "comment flagged but you may read it here at your own risk" button, or do a slashdot-like moderation (for newbies, slashdot keeps all comments, ye
    • Why are they censoring comments? That's just comments for chrissake!

      If the comments are so worthless that they're junimportant "just comments", then why bother having them at all?

  • Which comment would you block? Play the moderator role and take our quiz to see how your decisions compare to those of Guardian moderators

    In an opinion piece about what makes one a "feminazi": “Funny how so many journalists are female, and how many are feminists! A disproportionate number pollute journalism. Jusrt shows that men DO tend to do 'harder' jobs than keyboard bashing, while the technology that men designed and built is used to provide these harpies with a medium from which to spout their bi

    • The Guardian decisions seem quite reasonable to me. The censored comments are generally offensive, frequently assume the author's pet position with no backing, and really don't add much to the discussion.

  • "The Guardian Publishes Comment Abuse Stats, Invites Debate On Moderation" - by disabling commenting on the article in the link. I'm a regular on the Grauniad's website, and their censoring of comments has become ridiculous. I've have comments removed because I highlighted an author's hypocrisy, use of logical fallacies and general intellectual dishonesty.

  • I recently gave up visiting the Guardian website after being an avid reader since my student days in early 80s. The reason for this was an increase in click-bait articles combined with heavy handed and capricious moderators in the comments section.

  • by Stephen Chadfield ( 7971 ) on Wednesday April 13, 2016 @07:07AM (#51898367) Homepage

    My comment that people take screenshots of their Guardian comments and tweet them with the hashtag "#censoredbycif" was itself removed by the Guardian's moderators.

    http://www.chadfield.com/2015/... [chadfield.com]

    • When I get off my arse, I want to write a script that scrapes comments, and if it sees they've been removed, flag the cached record as such. That way I can see what's been deleted. It would be pretty easy to then tweet them automagically. In fact, that might be my task before the pub quiz tonight.

What is research but a blind date with knowledge? -- Will Harvey

Working...