[youreincorrect]

Exactly.

And, furthermore, the finding is akin to a form of survivorship bias. 'We found they stopped talking to each other and promoting hate online.' Okay, and what did they do next? Did they just stop hating people? Did they figure out a way to start meeting in person without Facebook to mediate the conversation? Did some of them become increasingly isolated that they decided to act out? Did some people who weren't members of the hate group happen to see this form of censorship and start to change their own minds? I'm not saying any of these outcomes are what happened or are inevitable, I'm wondering how they can call this real science when they don't even bother looking at the other effects outside of their water-is-wet conclusion.

[krapp]

>Okay, and what did they do next? Did they just stop hating people? Did they figure out a way to start meeting in person without Facebook to mediate the conversation? Did some of them become increasingly isolated that they decided to act out? Did some people who weren't members of the hate group happen to see this form of censorship and start to change their own minds?

Given that the target audience they were studying were users of the platforms, and the effect was the consumption and spread of hate speech within that population and on those platforms, I don't think that including n+ order effects across every other facet of society would have been necessary, or even possible.