2 ms·
There was an interesting point in a different post [1] in this topic. Essentially Twitter not only has a largely biased sample in terms of who uses it, but the
by somenameforme 5y ago
There was an interesting point in a different post [1] in this topic.
Essentially Twitter not only has a largely biased sample in terms of who uses it, but there's also an extreme bias in terms of who actually posts on it. About 22% of Americans (2019) use Twitter and, of that 22%, 10% are generating 80% of the posts. So Twitter is primarily driven by a bubble of 2.2% of Americans who skew younger and more liberal than the average. I would hypothesize (though it was not stated by the sources given) that the 2.2% who are posting like crazy also likely skew more radical than average as well.
It's quite dangerous because sites like Twitter can give people the imaginary perspective that they're interacting with hundreds of millions of people, when in reality they're generally communicating with an extremely unrepresentative and small sample of the world, or even of the "Western world". And now factor in the growth of tools, such as this, that add wide-range and spider-style ban/censor/muting of anybody who expresses a view you really dislike.
[1] - https://news.ycombinator.com/item?id=30838517 https://news.ycombinator.com/item?id=30838517