A Woman on Twitter is Abused Every 30 Seconds (fastcompany.com)

That shocking statistic comes from a study conducted by Amnesty International and AI software startup Element AI. From a report: In the study, called Troll Patrol, Amnesty International and Element AI looked at data from 288,000 tweets sent to 778 female politicians and journalists in the U.S. and U.K. in 2017. Using machine learning on the data, the group then extrapolated just how wide-ranging abuse toward women is on Twitter. The result: 1.1 million abusive or problematic tweets were sent to the women in the study during the year -- that's one abusive or problematic tweet every 30 seconds. And it's even worse for women of color -- and especially black women -- who were targeted more frequently than white women.

Comments

Machine learning

[AHuxley]

that criticism is trolling.

What does problematic mean?

[Anonymous Coward]

How many men are abused or sent problematic tweets on twitter.

Why limit research to one segment to create a self confirming headline?

Re:What does problematic mean?

[easyTree]

Who cares, only women and children matter. Men are to be sent into battle.

Re: What does problematic mean?

[phantomfive]

They really needed a control group here.

Re: What does problematic mean?

[Anonymous Coward]

That study is completely useless tripe by a dimwitted liberal trying to make some point that 'women have it hard'. You know what, yeah they do. So do men. So do straights. So do whites. So does everyone. These liberal whiney losers think they are so smart because they 'have shown how hard it is for women'. Complete imbeciles. I can name a single guy who gets abused 30 TIMES PER SECOND on Twitter. So that's 900 times as frequently, by 1/778th of the population. Let's see, simple math would say that it looks like men must have it 700,200 times as hard as women!?!?

Oh, that doesn't compute? You're right, it doesn't, and making such an ascertation would be irresponsible. But so is this study. Even without saying that 'this is soooo aweful', it has clearly been written to imply it as somehow lopsided against women and how bad all us terrible misogynists are by NOT MAKING ANY ATTEMP TO NORMALIZE WHAT THAT MEANS OR HOW IT RELATES TO THE HARDSHIPS OF ANY OTHER GROUP. That omission is designed to elicit anger. So until those irresponsible f-wads fix their study and all the articles that devolved from it, I will stand with my fact based analytical response that it must be 700,200 times worse for men based on my sampling, and I didn't have to be some liberal whiney weeney spending 6 months to come up with that.

Jesus, and they keep blaming conservatives, but this BS that is designed to create strife while trying to feign innocence flows ever hour of every day from liberals. You know what? It's working, strife has been down and conservatives either have liberals because we can see right through this BS, or because we know a large percentage of your are so stupid you actually believe this BS is real and that women somehow have it so much more awful than men.

Any woman who is on Twitter has it a thousand times better in every respect than quote not on Twitter. Holy f, go see how women in third world countries are treated if you want to talk about how bad women have it. In virtually every first world country women have it way better. You see people tripping over themselves here to come and and a woman being assaulted or in distress while meet get hammered and other men then jump in to hammer the6n further.

This sort of crapola is why conservatives really are starting to *really* hate liberals.

Stop it.

Re: What does problematic mean?

[karlandtanya]

Folks, this ain't science; if you measure it by that yardstick it will come up short.

To answer your question, it's because the popular narrative is that women are victims, passive, and must be protected while not-women are villains, aggressive, and must be punished. Of course, repeatedly recognizing that your sample set ' 778 women politicians and journalists with an active, non-protected Twitter account, with fewer than 1 million followers' is flawed doesn't make it any less so. The sample isn't chosen to test a hypothesis; the goal here is to tell a story.

Element AI has provided a sample of their work to Amnesty in exchange for some goodwill. Decorating their product in the style of a scientific study is similarly insincere.

As an ad for Element AI it's valid: We can generate buzz for your issue.
As a political statement from Amnesty International, it's also valid: We need more civil and more enforcable social norms in social media.

As social science it's crap, but that's neither Element AI's or Amnesty Internal's job.

Re:What does problematic mean?

[sfcat]

Why limit research to one segment to create a self confirming headline?

To answer your question: because you can't study everything at the same time without making assumptions you can't justify yet. This happens all the time in social sciences. Before you can safely lump things together you have to study them separately.

I was going to mod you down for this but I will comment instead. This isn't science. If someone thinks it is, then they need to turn in their science card at the library. Anytime you setup an experiment, you have a control group. The amount of anything experienced by one group only has meaning when compared to the amount experienced by another group(s) which usually includes a control group. In this case, probably a group of men (or at least non-women). If it came out that the control group is attacked by someone on twitter every second what conclusion would you draw from those two data sets. If instead it was once every minute that a person in the control group was abused, would your conclusions be different? Add to that that possibly there is a gender imbalance in numbers of interactions per capita between the two groups and this one bit of research is meaningless. And it's stuff like this that causes "social scientists" to be considered junk science and for their entire field to be put into quotes by other scientists. Because an experiment without a control group isn't an experiment. Its somewhere between bad science and propaganda.

Re:What does problematic mean?

[umghhh]

I wonder what is this social science, you talk about? I mean science is a way to find explanations for things based on facts and reason. According to wikipedia and my dictionary the word comes from Latin word meaning 'knowledge'. How does social science with significant 'social justice' element (i.e. bias) fit into this? If even particle physics does exclude people based on their (male) gender then social science cannot really be better.

As for your claims of them social 'scientists' not having done full analysis just yet and hushing with their significant results. This is just plain wrong - in modern times a claim of females being oppressed has such political weight that no serious scientist should have claims based on partial data. Then there is this other thing: besides ex.boyfriends having the urge to call their former better parts names - how many of the bad words (assuming they are really bad and not something like: 'this was a mistake because...' or 'you looked better in another rock' etc) were written by females? This again causes serious problem in our culture sliding ever faster into matriarchat.

Yet another thought I had - according to modern social science the gender thing is nothing more than a cultural construct. How does that work here - are these abused females sure about their identity and how many of the abusers can be culturally identified as males and females and how does that look like in a comparison between all combinations?

Besides all this, I do not think we can have facts and reason based arguments about such subjects so I apologize and commit for searching of reeducation camp in the vicinity, being white old male etc.

Re:What does problematic mean?

[houghi]

They are not studying everything at the same time. They study something very specific. They study "Abuse on twitter". This is not so different to "Xmas messages on twitter".

So you first see how many messages are send. That you compare to the total amount send. This both in total and per individual. You will see that some get none and some get a lot.
You can even analyze from how many people it is from. You will see some will send nothing and some will send a lot. Some will even send something to themselves.

Once you have that information, you can look at different factors. e.g. country of where the message comes from and is send to. You can also loom at age, gender, activity in general per account and a lot more things.

The focus could be a conclusion of a genral research. You should not START with a focus group, because that will skew the results.

As you stated,the best social studies tend to take big trends and parse then into pieces. If you start with only women, you start with peices already.

They should have gone with "Abuse on Twitter" instead they went with "Abuse on women that use twitter". This is just as horrible as looking for "Criminal behaviour among black males" or "Greed among the Jewish population" or "Patriachial behaviour among fat bold white males". Because you are clearly steering the investigation.

And let us look a bit closer. From the website: "778 women politicians and journalists". So not only are these women, these are women in jobs that are in aprofession that will get controversial reactions, no matter what.

e.g. if they have an opinions on e.g. birth control, no matter their side, somebody will be against their opinion. And that is what people will tweet. So that number of one every 30 seconds suddenly sounds very low. It is ONLY one every 30 second? That's it?

Re:What does problematic mean?

[roman_mir]

Men have the worst deal. Men don't want to go and die in wars. Most women with toddlers wanted those toddlers. I am sure if given a choice to go to war or to look after a toddler most men (most normal people) would choose the latter.

Re:What does problematic mean?

[DaveV1.0]

Oh, I didn't realize that caring for a toddler was such a hardship that comparing it to running the risk of being killed or permanently maimed was a valid comparison. Are you a troll or just an asshole?

Re:What does problematic mean?

[DaveV1.0]

You mean a group of people with a specific characteristic are subject to attack based that same characteristic? Shocking! It is all just grievance studies, in this case, they have extrapolated high profile people to the whole of twitter to come up with a fake statistic. This is called sample selection bias.

Re:What does problematic mean?

[dcw3]

As a veteran and 7 year single dad, I'll answer for you. Taking care of a toddler is cake compared to war. Got it?

Didn't measure/compare against abuse rate for Men

[Anonymous Coward]

Without a meaningful baseline to compare against, what does the stat mean? (from what I've read, they didn't bother measuring the abuse frequency against male tweeters - sort of the same as when it's mentioned the appalling statistic that 10,000 women committed suicide in 2016 whilst deliberately neglecting to mention that 35,000 men committed suicide the same year)

Men?

[Anonymous Coward]

Not a peep about how many abusive tweets were sent to men, using the same criteria.

Why would that be, exactly? Is it because women are the weaker sex, and hence need to be protected from this sort of thing, whereas men are stronger and hence don't need such protection?

Or is it that nobody gives a shit about men, whether they are stronger or not?

Or maybe men get WAY MORE abuse on twitter, but pointing that out won't substantiate the desired narrative?

Lies, damn lies, and statistics.

[microbox]

Typical "intersectional" reporting. What is the base rate of abuse? Always be suspicious of people who quote absolute statistics. If you look at the research on this, you find that men are abused more than women, and that when a woman receives abuse, it is more likely to be from another woman. But the facts don't suit the narrative of the poor innocent woman being persecuted by the evil mens. PATHETIC.

Re:Lies, damn lies, and statistics.

[Shaitan]

What are they even defining as abuse?

Re:Lies, damn lies, and statistics.

[taustin]

"Any post made to a woman that we don't like, or that conflicts with our political agenda."

Duh.

Re:Didn't measure/compare against abuse rate for M

[Shaitan]

And what constitutes "abused"

I think I found her husband

[rsilvergun]

right here.

Is this Slashdot or SJWdot?

[oogoliegoogolie]

I thought this was news for nerds,news that matters.

Amnesty International-- really?

[Harvey Manfrenjenson]

The disturbing part isn't that somebody published this absolutely idiotic study. (I won't go into why it is idiotic, since literally every other post has pointed out one or more serious flaws in the study-- rarely have I ever seen anything ripped to shreds quite this thoroughly). No, the disturbing part is that Amnesty International was involved. THIS is what Amnesty International is doing these days?

I mean I'm not an expert, but I always thought Amnesty International was one of the most respected of all human-rights organizations-- the sort of organization I would donate money to. Not any more.

Re:Didn't measure/compare against abuse rate for M

[Anonymous Coward]

And what constitutes "abused"

Saying anything I don’t like.

Re:How is someone "abused" by a tweet?

[Mal-2]

I know where you live, bitch, and you have to sleep some time. https://www.amazonsellerslawye...

You can fit plenty of intimidation into 280 characters.

What the hell is this shit?

[NoSleepDemon]

What the absolute fuck has this site come to? If someone went back in time and told cmdrtaco what was to become of his beloved tech news aggregator after his departure, I wonder if he would have believed them? This place might as well not exist anymore; it is nothing more than a cash cow to grab ad revenue by posting the most obviously inflammatory bollocks its moderators can think of. Whoever posted this must have cum in their pants at just the mere thought of all those ad impressions. Fuck me, what a travesty. I'd say the new owners ought to be ashamed of themselves but they either don't know what shame is, or it's something they enjoy feeling. I've half a feeling that they even own those dopey nazi ascii posting bots as well; anything to make a dollar, the dirty cunts.

Matthew 11:8

[Hognoxious]

This. Without any baseline to compare against the statistic quoted is meaningless.

Re: A Nerd on Slashdot is

[Ol Olsoc]

It also very much depends on how you define attack. I would rather like to see the precise definition before rushing to any judgement.

I think you'll find that opinions run the gamut from actual abuse, to disagreement.

These claims often trend to the silly. My wife is asked out all the time on line. In today's world, that is sexual harassment. She just giggles and tells the guy he's not man enough, but many women need intensive counselling for the grave damage done.