We built a data set of 45 million comments on news articles on the Huffington Post website between January 2013 and February 2015.
I am no expert but I feel like this is a really bad data set choice for this study.
we built a dataset of three of my comments and found that…
It’s just a bad data set for basically anything
Yup, comments on news articles are pure cancer. Comments about news articles can be decent though, but they need to be hosted elsewhere.
A persona allows people who otherwise wouldn’t to experiment with being magnanimous and making/admitting mistakes(some more cautious IRL, others incapable of backing down).
OTOH, there’s always people who play Paladins on tabletop, and the mitigating factor of the Block button - most who don’t want to aren’t forced to engage with the worst of us, and blocking someone who knows you IRL has a more complicated cost/benefit calculation. This is one thing I feel cancel culture and the younger generations get right; Screw the other consequences when not blocking that shitty uncle, boss, teacher, coworker, celebrity, whoever, is letting them monopolize some of your personal time and mental energy.
How dare anyone impugn the integrity of Weedlord Bonerhitler69? The man stands as a colossus of virtue.
am still sour [._.]
Interesting. But the article headline is misleading. The article states that the biggest difference was between volatile anonymity where people could make arbitrarily many accounts, and stable pseudonyms, where a ban cannot easily be evaded. Stable pseudonyms are a lot better as the article states.
Between stable pseudonyms and real names, the difference is smaller, as stated in the article. Real names make it only slightly worse.
Or in other words, effective bans work best for moderating a community. Shocking news I tell you.