An experiment in fighting racism on Twitter
washingtonpost.com
washingtonpost.com
It's a shame they over-hyped the title and made it all clickbaity. The post itself clearly states that they are not autonomous and therefore not programmed. Possibly not helped by the use of the word "bot" over and over. Am I right in saying "bot" implies autonomous?
That aside, interesting short read with a fairly intuative result. Good to see it backed up with actual research. Though it doesn't mention how many accounts were targetted during the research.
Is there potential a way of analysing messages for racism in the same way some do sentiment analysis? Is it the same sort of problem? Not even a beginner in the field, let alone an expert!
In Twitter's case, follower graphs would also probably be a good indicator, at least for people who follow less than 200 accounts. Not to mention that a lot of these people overtly say in their twitter bios that they're part of the frog police or whatever.
Twitter could nuke Alt-right Twitter from orbit pretty easily. Then they'd just go run off and use some "actual free speech" twitter clone and fester in their own space. Just like they do when reddit/4chan cleans house.
I just had a similar chain of thought at a story of a students union wanting to ban certain newspapers from the campus. My opinion is that, in that instance, the union should focus on the people who buy those newspapers. Why do they buy them? Do they agree with the messages? Or perhaps they are believing what they see without critical thought? There is then either an ideological debate or perhaps critical thinking education that can be undertaken.
Hiding the message of an organisation or individual seems morally wrong to me. I'd much rather see the opinions published and shunned rather than outright censored. In the Twitter example, I'd like to see the classifications used to help understand the viewpoint of others or provide further information prior to posting, rather than to censor / ban individuals or groups.
Not a fully formed idea, for which I apologise. My gut (and it is only my gut) tells me there is a useful aid there.
At it's core, twitter is an assemblage of communities. Twitter can simply decide that, no, it doesn't want neo-nazis on its website. These people are then free to go make their own websites to hang out on.
Websites like Digg or Something Awful, despite also not being about single topics, were never sitting around thinking "Oh, I want to ban this guy that goes around calling black people monkeys, but free speech". Even 4chan would ban people for that sort of dumb stuff. I feel like this argument only became prevalent after reddit became big.
There is an argument to be made that reddit's philosophy on this point single-handedly made all "community" websites worse, because it stigmatized the notion of "drawing a line in the sand", and implied that all websites have to accept everyone.
I think in the newspaper example, Twitter is a newspaper, not the campus. People can make new websites easily, just like they can make new newspapers. Twitter is under no obligation to "publish" stuff if they don't want to. The internet is the campus. No DNS blocking, just banning users from some websites.
And, honestly, the amount of people who care about the "free speech" issue dwarfs the amount of people who see Twitter as "the place to get trolled at". It doesn't seem like good branding to be at the same level as 8chan on this point.
Though, to your point, on a more general level, figuring out what's going on in the audience for these sorts of messages is important.
The problem with the alt-right is that it is an idea. Ideas are very hard to kill by censorship and when you do let them "fester" they get more extreme and radical. This is in part how we arrived at the alt-right becoming what it is today.
We should want the alt-right out in public in spaces where there are checks and balances. There, it can be killed by moderation (in both senses of the word, organic and inorganic) and dilution.
If you push the alt-right to exclusively use their own walled gardens, to concentrate in their own unmoderated, subverted echo chambers, that's when you should be very terrified because that's when you birth actually scary groups.
There will always be unpopular opinions on some subject. No matter how much proof you have, there will be flat earthers and creationists. No matter how much you promote equality, there will be bigots. But there is also always a continuum on the spectrum of these affiliations - some people may be mildly xenophobic and have valid concerns on immigration or have ignorant/antiquated views on gender equality, for example.
When you take a hard-line stance on these opinions on the hope of promoting one ideal (no matter how noble) you always run the risk of also promoting the other through polarisation.
This idea that the 'right' is a problem is in itself part of the problem.
There is plenty of harassment from those that claim they are more tolerant than the rest of us.
I do agree with your echo chamber point though -- one needs only to spend a few minutes on Mother Jones's comment section to read all manner of hate against those who have different opinions from their own.
You are exactly correct -- the public sphere ought to welcome all viewpoints as sunlight is ultimately the best disinfectant.
One of the reasons Hillary lost the election is that her campaign seemed to be in a bubble -- holding valid concerns of many groups in contempt while surrounding itself with voices that spent more time self congratulating or feeling intellectually superior whilst many Americans (such as in the Rust Belt) found themselves marginalized. The Obama campaign by contrast actually reached out to the so-called deplorables.
Vitriol is clearly a bad measure when subjective interpretation rules are becoming official policy. Even Twitter now has special treatment for particular groups enshrined in its policy. You can be as vitriolic as you want, as long as it's against the "correct" group.
The Alt Right is anti-equalitarian. It rejects the idea of equality for the same reason it rejects the ideas of unicorns and leprechauns, noting that human equality does not exist in any observable scientific, legal, material, intellectual, sexual, or spiritual form.
From http://voxday.blogspot.com/2016/08/what-alt-right-is.html
Good that there's a number of items on the list I agree with. Some common ground.
The general answer to the apparent contradiction is generally found in the other points and implied in #15: complete separation, e.g. #11 "The Alt Right understands that diversity + proximity = war." No proximity, no war of that sort e.g. embedded in those laws you refer to, and therefore no problems with #14 of those sorts.
Also, I should make clear this is Vox Day's first cut of a statement of what the Alt Right is, but he thinks it's solid enough to get it translated into quite a few languages now. But the Alt Right is still seriously inchoate.
"#11 "The Alt Right understands that diversity + proximity = war."
That needs some explanation. What does that mean exactly?
It means if you put enough people of different identities in close proximity, the inevitable result is "war". Maybe not Yugoslavia type war, but warfare of one type or another, including "simple" economic warfare, which was one of the causes of The War Between the States, and had been a frequent problem for the young Republic prior to that, but it also happens on very small scales, see for example the various non-native merchant classes in so many countries, and that doesn't require large numbers on both sides.
There's a social scientist who's research said much the same thing, who published his data early but otherwise sat on his results for more than a decade, Robert Putnam. As Wikipedia puts it, "His conclusion based on over 40 cases and 30,000 people within the United States is that, other things being equal, more diversity in a community is associated with less trust both between and within ethnic groups." (emphasis in original: https://en.wikipedia.org/wiki/Robert_D._Putnam#Diversity_and...).
"The Alt Right believes Western civilization is the pinnacle of human achievement and supports its three foundational pillars: Christianity, the European nations, and the Graeco-Roman legacy."
Smacks of racism... But we all need our selective foundation myths.
"5. ...It supports all nationalisms and the right of all nations to exist, homogeneous and unadulterated by foreign invasion and immigration."
So does the Alt-Right support the right of self-determination and national autonomy for the Palestinians in Israeli-Occupied Territories?
Does the Alt-Right aspire to the German National-Socialists of the 20's-40's?
"7. The Alt Right is anti-equalitarian. It rejects the idea of equality..."
So not all people should be equal before the law?
"scientodific" - is that even a word?
"10. The Alt Right is opposed to the rule or domination of any native ethnic group by another, particularly in the sovereign homelands of the dominated peoples."
I think Native Americans would agree with you here.
"14. The Alt Right believes we must secure the existence of white people and a future for white children."
And then the overt racism.
"15. ...human sub-species..." I didn't realize there was such a thing. Last I checked DNA is DNA and sperm fertilizes egg.
This whole things is racist. It is small-minded, backwards thinking, isolationist, static and reactionary.
These ideas are further an affront to the Enlightenment ideals that are one of the true cornerstone of western civilization, whose roots go deeper than Greece and Rome, and the progress that we all benefit from - https://en.wikipedia.org/wiki/Age_of_Enlightenment
Get outta Dodge yo.
Yes, but there has to be physical separation. So one way or another, the "occupied" has to go. That's explicitly covered in #10, and #15 says at the end "Every race, nation, people, and human sub-species ... possesses the sovereign right to dwell unmolested in the native culture it prefers."
Does the Alt-Right aspire to the German National-Socialists of the 20's-40's?
No, if for no other reason than that they lost ^_^! The Alt Right is about winning, and the NASDP weren't winners, from the individual level (see how many biographies, granted, biased, say pretty early on "a failed X", starting with Hitler), all the way to the international level.
So not all people should be equal before the law?
That's pretty clear from many of the other points. E.g. a person from "race, nation, people, [or] human sub-species" who's not in one of their nations would not have the same rights as natives of the nation.
"scientodific" - is that even a word?
It's a supremely infelicitous neologism coined by Vox Day that I don't see anyone else picking up because it's so bad. But it's defined in the next sentence of the point, so, eh.
I think Native Americans would agree with you here.
Which just so happen to include Vox Day ^_^ (long suspected by he and his siblings, then recently confirmed by genetic testing; he's also of Mexican descent. No doubt part of why he's part of the Alt West, not the Alt White, although there's also mere details like Christianity differentiating the two subsets of the Alt Right as he's defined their taxonomy, as noted by many including Hitler, Jesus was a Jew after all).
I think the "human sub-species" being referred to in #15 pertains to the recent discoveries WRT to Neanderthal intermixing and the like, but this is not an topic I follow. Ah, per Wikipedia, and this is following earlier research first published in 2010 (https://en.wikipedia.org/wiki/Neanderthal):
"In 2016, research indicated that modern humans had three distinct interbreeding events with Neanderthals: the first encounter involved the ancestors of all non-African modern humans, probably soon after leaving Africa; the second, after the ancestral Melanesian group had branched off (and subsequently had a unique breeding event with Denisovans); and the third, involving the ancestors of East Asians only.
In short, only members of the Negroid race are pure homo sapiens ^_^.
Last I checked DNA is DNA and sperm fertilizes egg.
When it comes to defining species and sub-species, not exactly; per Wikipedia (https://en.wikipedia.org/wiki/Species and https://en.wikipedia.org/wiki/Subspecies):
A species is often defined as the largest group of organisms in which two individuals are capable of reproducing fertile offspring, typically using sexual reproduction.
A common way to decide is that organisms belonging to different subspecies of the same species are capable of interbreeding and producing fertile offspring, but they do not usually interbreed in nature due to geographic isolation, sexual selection, or other factors.
1. Alt Right is right-wing. No lefties!
2. Alt Right so edge, not lame. Cucks and cons so lame.
3. Alt Right gonna win. Not gonna lose.
4. Europe good, church good, white people good. History good too.
5. Freedom for everybody!
6. Stay home and don't touch the foreign poop.
7. Equality bad, equality isn't even.
8. Science good. Made-up shit bad.
9. Everybody votes they boys.
10. Small crew can't tell big crowd what to do. Not for long, anyhow.
11. MS-13 and Latin Kings gonna kill Crips and Bloods. And vice-versa. Keep them apart.
12. Alt Right don't give no fucks.
13. Free trade bad. No jobs, no money!
14. White people good.
15. Leave all them Red, Yellow, Brown, and Black people alone.
16. War bad, genocide bad too.
TL;DR: Everyone should be left by they own damn selves
The UK's Brexit vote was swung by the fact that some of our newspapers have lied about the EU for decades. The list of lies is breathtaking. They literally just make shit up to further their cause.
The Left doesn't lie and manipulate on an industrial scale like this. The Left sometimes gets facts wrong, but there is no industrial fake-news network supporting left-leaning ideas. (And no - the mainstream media are not that network. The MSM are primarily pro-corporate and pro-establishment, not left-leaning.)
Once you start lying and manipulating professionally and commercially, notions of free speech no longer apply. Industrial dishonesty is toxic to democracy. It's also an easy and cheap tool of influence for hostile nation states who want to undermine shared values.
https://www.youtube.com/watch?v=Drn0Lg5gIRc
That's one of numerous examples of selective editing by CNN.
> The Left doesn't lie and manipulate on an industrial scale like this. The Left sometimes gets facts wrong, but there is no industrial fake-news network supporting left-leaning ideas. (And no - the mainstream media are not that network. The MSM are primarily pro-corporate and pro-establishment, not left-leaning.)
Even if that is true now (and it's a very dubious claim), you don't have to go back very far in history to see the left bending over backwards to justify the socialist utopia Stalin was building in the USSR.
If you believe in truth, if you believe in honesty, stop making excuses when the liars are on your side. Stop using the no true Scotsman fallacy to excuse liars. Dishonesty is used by human beings of all classes and political stripes. It always has and it always will be.
Let's call it out when we see it, leaving partisanship out if it as much as we can. This could be an issue that unites us, rather than splits us.
Something that makes this increasingly difficult is that we're having difficulty agreeing on shared truth. I think it's right to increase scrutiny on sources that are proving themselves untrustworthy, but we need to start trusting (and verifying) somewhere.
One things that separates the Alt Right from in the US the moribund and dying movement conservatives and libertarians is that we're willing to use any weapon the Left uses. That is, one thing that distinguishes us is that we refuse to play the game by the rules set for us by the other side, rules that ensure we lose.
So while we may find lies in the context of public discourse distasteful, they're certainly on the menu of tactics, although they're frankly weak sauce and hardly needed, like every other one of the Left's. The latter being a legitimate reason for the Left to be totally freaking out right now.
As for "unites or splits us", you're living in the past, we're split, ever increasingly during my entire lifetime (and I turn 56 RSN), can you honestly say there's any potential for this trend to not continue?
Put yourself in the shoes of the Right; what is the Left offering that would move us back "together" in some sense?
I think you are wrong about that. There's also the issue of the underdog not having the institutional control (media, fact-source institutions) to really build an effective counter PR scheme.
But I think there is a fair amount of evidence from this election that the increased velocity of idea transmission on media like Twitter made it impossible to dilute. That's one of the factors in information spread models, and it seems relevant here.
The alt-right used to be in their own walled garden echo chambers: Breitbart, Stormer etc. Now Steve Bannon is two months from being chief-of-staff in the Whitehouse.
So I guess that didn't work out so well.
ADDED: he is formally "chief strategist and senior counselor".
I believe allowing these people on your community site gives them the space they need to become more extreme and radical. The standard you walk past is the standard you accept, and it's always struck me as very convenient that Reddit's minimal censorship policy magically aligns with minimisation of operating costs.
I think this is not only wrong, but troubling. tolerating actions on the street is not the same as tolerating speech in a private forum.
I thought the point of twitter is that everything is public?
I never dreamed I'd see the day when electing a president was called radical.
It makes it easier to paint a rosy picture of your cause when the previous stuff you said has vanished from public view. It is very possible that censorship of ISIS content on Twitter (I'll take an extreme example) actually has made it easier for them to spread. You are both able to rewrite history and harness whatever in a human brain likes things which are forbidden (aggressively used by musicians and entertainers for decades in the West to market to a younger audience.)
The biggest weapon against people with really bad ideas is definitely an undeletable history. In Europe, they no longer have that tool.
Funnily enough, in the UK we're about to pass into law a ridiculous bill that forces ISPs to store the last 12 months of browsing history for every user. This is amongst other wide sweeping powers such as being able to ask companies to do the impossible (Apple to break into a phone). Definitely something I personally will be finding out the finer details of.
Had they been present, they could've weakened the support a bit on the more moderate ones. At the least, whoever they propped up would reflect less-extreme views of their supporters. Isolation only makes bad things worse.
I feel like I'm justified in being much more terrified when the white nationalists are out in the open, empowered to inflict violence on people because they're confident country has their back.
Like, you can be terrified on an abstract level because you think the anti-nazi politics aren't being performed optimally, we'll be over here fearing for our lives.
A politician was assassinated in the UK suspected (read: highly likely) over Brexit. Don't assume things can't get even worse is what I'm warning...
Maybe they'll say something like "yeah we should kill politician xyz!!!!!"
If it's in public, chances are, more normal people (who happened to also be part of the discussion) will recoil and you can hope that person will realize that they overstepped something and dial back their thoughts/ideas.
In a private space, where the discussion is framed by the founders and newcomers are joined (possibly only after being "vetted") it's probably more likely that after a suggestion like "let's kill xyz" will be escalated with something like "great idea - I know person A who sells guns on the side.
And suddenly that person who possibly could have been talked out of doing something tragic has instead been empowered to act.
I have faith that most of us are not malicious, but are still susceptible to the animal in us and all the instincts that come with that...
But there's a lot of nuance between hiding them, and not giving them a platform. The points of view of the President-Elect aren't in any danger of disappearing from view, they'll be published and discussed in the media regardless, and there's been plenty of white-nationalist websites with discussion forums for everybody in the alt-right to present their message.
Another area where one might find nuance is between expressing viewpoints, and harassment and hate speech. It's perfectly possible to express oneself and one's politics respectfully. I don't think that's what anyone is talking about when there's talk about twitter needing to clean up their platform. Contrasting viewpoints are not the point of contention, it's whether people should be supported when they go out of their way to hurt and abuse and intimidate other human beings.
Should twitter really feel obligated, morally or on any level, to sacrifice the well-being of the victims of slurs and harassment, only because of the (very tenuous and abstract, in this case) danger of censorship?
Though with international platforms jurisdiction with law enforcement can be tricky, I'm sure I remember cases brought to court successfully off the back of Twitter abuse.
Not disagreeing with your points entirely though. Perhaps there is middle ground in the platform providing analytics of this kind to law enforcement?
On one level there's the immediate individual occurence of harassment where I agree law enforcement should step in where applicable, instead of putting that responsibility on the platform owner (though it doesn't seem like law enforcement is typically equipped or motivated to do so, dedicated analytics seem like a good step towards addressing that). But I don't think that needs to preclude the platform taking action as well, simply to improve their product for the vast, vast majority of users who don't get off on harassing marginalized people.
The other level is the macro scale effect of being the platform that's cool with white supremacists chasing other people away. Independent of your motivations for having a hands-off approach to moderation, you're gonna be emboldening them, making them feel welcome and comfortable and consequently growing a toxic user base. In turn, people who are particularly subjected or vulnerable to that sort of harassment will act in their self-interest and get the fuck out. "No moderation" isn't a neutral option, and law enforcement isn't going to take care of community management for you, so I think this is a strong argument for the responsibility of the platform to step in.
I'm not sure if the royal mail comparison is really applicable. I guess you can frame every platform as some sort of public utility, but I think that analogy works better for lower-level sorta things like OS developers or CDNs. In particular I think the argument doesn't do justice to the role of twitter as a social space. Maybe a better comparison would be whether you'd expect a bar to continue to provide service to patrons with a history of starting fights with other customers? You'd also expect law enforcement to arrest people for assault, but I think it'd be reasonable that they wouldn't be let back in after trashing the place anyway.
What would you say to banning and retaining the history. Tweets all appended with "user banned for x"? Like a bar with photos of banned patrons.
Comparing across the panels of Figure 4 shows the decay in the effect of Treatment C over time. Although the effect remains statistically significant, the coefficient decreases steadily. In Panel A, the point estimate of -.27 indicates that the daily rate of the use of the word “_” decreased by .27 more among subjects in Treatment condition C than among subjects in the control condition. This average treatment effect for condition C decreased in magnitude to -.17 in Panel B and -.11 in Panel C. I collected data for 2 months, but these results are not shown because none of the treatments are significant.
...
Encouragingly, these effects persisted over time, for the first month under study, although not for two months. Also, the effect was significant at p < .05 in the two week time period, but it was only significant at p < .10 in the one week and one month time periods.
Wow! The subjects cut back on their slurs for the next month (and what is the practical difference in .27 fewer daily slurs compared to the control group?)....and then they started doing it again.
http://kmunger.github.io/pdfs/Twitter_harassment_final.pdf
Can't help but wonder how you get a Washington Post article out of that. It's a good graduate-level semester paper. What connections does this guy have?
The harassment was some pretty serious shit (making a twitter account of the victim's dead father, then insulting through this). But one day the attacker realized how awful the thing they did was, and ended up apologizing profusely to her.
There was also that study about how people's views on gay marriage evolved substantially after just a couple conversations with a gay couple, who would say "Hey, why can't we get married?". Though there was some statistical issues with that IIRC....
I have a hard time saying "these are proof that yelling at trolls works" because there are good and bad ways of communicating the point, but it sure feels like direct confrontation works on some level.
Though I don't know how well this stuff works against people from /pol/, if only because most of them don't actually care about the issues, just yelling at people.
[1]: https://www.thisamericanlife.org/radio-archives/episode/545/... (Act 1)
I disapprove of what you say, but I will defend to the death your right to say it. - https://en.wikipedia.org/wiki/Evelyn_Beatrice_Hall (commonly misattributed to Voltaire)
It's also easy to see an automated approach leading to hilarious results with say tweets from African-American rappers.
Finally, maybe we can move past this "might offend someone" stance that offers 0% nuance?
Or that in the past insults differed therefore we should ignore today's.
Or that hypothetically something hilarious might happen so don't call out trolling harassment.
BUT it would be unethical if it was used, for example, to create the impression that something is the prevalent or mainstream, or the most popular opinion when it isnt. propaganda tactics
End racism with software? Bold claim, and sign me up, even if it means manipulating racists out of their bigoted attitudes. I honestly can't see a down side to that.
Which, in turn, will make Twitter an endless bot flamewar.
It'll stop working the moment anybody notices.
In fact, its effects will be inverse from that point.
It's essentially spam.
I think Charles Stross tried to warn us about this. Get ready for weaponized memetics.
So, by all means, consider breaking the rules. And then don't, please.
Maybe it was only clickbait but they clearly suggested automation.
But now on second read I see it's ambiguous and the author seems to be politics student so maybe you are right after all that he really did it all by hand.
> Perhaps he just wanted something short and that sounds better than "fake user accounts".
I think "accounts" would be fine. Or "sockpuppets" if one wants to be fancy :)
Yes, but why did you pick a temporary one-man experiment as your hill to die on rather than a persistent, large-scale manipulator like Facebook?
If you're actually concerned about the ethics of manipulation, why didn't you say anything in any of the several HN submissions about Facebook's fake news issues?
It is true that there are larger manipulators of public perception - look at Hollywood - and I do think the same concerns should be leveled at them (eg. US military sponsorship and support for certain pro-US movies, product placements, etc.).
Ah yes of course - racism can only be carried out by white men after all.
What a pile of rubbish. If you take offense at words, things are going to get progressively worse for you as more and more automated bots enter the scene.
The answer of course, is for people to stop taking offense.
Do these people get offended at spam emails suggesting they need viagra or a fuckbuddy? What's the point?
...because he wanted to test the in-group out-group effect. Furthermore given the metric he was using...
Honestly, did you even read the article?
If a statement is absolutely fine if the person saying it is black, but absolutely offensive and racist if the person saying it is white, then it's the person taking offense who is the racist one...
2) No.
No one said that. Just that racism is an oppression where whites are the top caste. Blacks can be racist against blacks too, for example internalized racism.
However, blacks being racist against whites is an odd statement, like homeless people being classist against the rich. Race is pseudoscientific classification of people to justify entire economic systems like slavery, and still lives on like awful backwards compatibility.
> Do these people get offended at spam emails suggesting they need viagra or a fuckbuddy?
Companies like twitter are far more aggressive at fighting spam than physical threats. To the point where people simply reported abusers as spammers at one point. That's why no one bought them.
Oops. Maybe they should've been less arrogant to their embattled userbase; so-called "social justice warriors" turned out to be normal people just trying to help them make money. What happens social media companies refuse to learn things like sociology. http://www.businessinsider.com/disney-twitter-acquisition-tr...
I get what you're saying, but this just isn't a very pragmatic perspective. It's clear that white people have major hegemonic advantages over minority groups, but at the end of the day, the left (of which I am a part) needs to acknowledge and reject the damaging effect this kind of rhetoric has on racial discourse. As an extreme example, a statement like "fuck white people", is clearly racist regardless of who is saying it. It is difficult to fight racism if we give casual observers ammo to reject the left as hypocrites. This doesn't mean we have to pretend that the impact of racism affects all groups equally, but in order to see progress we must condemn hateful speech in all forms instead of dismissing it just because the speaker hails from a disadvantaged group.
The trick is understanding who holds the power in the situation.
And it indeed would be classist. You are neglecting this because you think that homeless people can not have any influence.
But consider this, group of homeless people ambush and attack people they consider rich. Would you still think that it is acceptable?
If we want to root out some behaviour then we can not tolerate it in any form.
Maybe there are automated ways around the problem but it's tricky balancing open discourse with automatic banning of abusive accounts since any kind of automated system will invariably stifle legitimate discussions. The word "nigger" when used in a discussion with the proper historical or even comedic context can be benign but I doubt any automated system will be able to make the distinction since that would require some kind of understanding of history and comedy. For example, https://www.youtube.com/watch?v=VOwjtNEoRYg
Happens in systems with Full Real Name policies as well.
IMO Full Real Name policies only deter people who 1. Have something to lose 2. Have a dissenting view 3. Does use a fake account with a Fake Real Name.
I.e. IMO a Full Real Name policy alone just sets us up for an echo chamber situation.
I have never conducted a study which involved actively intervening (and would like to know how IRB approval was obtained to do so), but the number of inferences which must be made are difficult enough when simply observing user activity. I would think running manipulation checks for the user's perception of the intervening profile would be nearly impossible.
Ultimately what is most difficult about Twitter studies is that the data collected is rarely available for inspection by anyone other than the researcher collecting it. When so many subjective decisions must be made in analysis this is particularly troublesome.
Right, because white men are the only people are the only ones who can be racist. If you have to rely on a criterion like this, that's a very good sign that the method you use is not objective.
Besides that, I really like the reply the author chose, the message and its phrasing seem very effective, while staying respectful towards the tweeter.
> It was essential to keep the race and gender of the subjects constant to test my central question: How would reactions to my sanctioning message change based on the race of the bot sending the message?
Why they chose people who were white shouldn't really matter as their conclusion was not "white men are racists". It was just something that had to be kept constant and grouping on race is not something unfamiliar for researchers, especially since in this case the whole experiment was based on the races of the sender and receiver. (Also note that it's much easier to distinguish between black and white rather than if they made the profile pictures different shades on the Fitzpatrick Scale)
So for instance, no, race is not a biologically meaningful signifier.