You (Yes, You) Could Be an Internet Troll
Ah, internet trolls. The type of people who make you wonder how human beings haven't been karmically smote off the face of the earth by now. These nasty folks lurk in comments sections and try to make other people feel bad, stupid, ugly or all three. But, mercifully, this kind of person is a rarity—right?
According to recent research from Stanford University and Cornell University, all of us are capable of being a troll, given the right circumstances.
Why do we troll?
The research team asked participants to take either an easy or a hard test, answer some questions about their post-test mood and then comment on an online news article. (The people who took the harder test were in a worse mood than the people who took the easier one.) Some of the participants saw an article with three pre-existing neutral comments. Some saw an article with three troll-y comments. They were required to leave at least one comment but could also respond to, up-vote and down-vote other comments.
Thirty-five percent of the participants who took the easy test and saw the neutral comments posted troll comments of their own. But that number jumped to 50 percent for the participants who took the hard test and saw the troll comments, suggesting that our mood and the commenting atmosphere can cause us to act out online.
Another experiment looked at more than 26 million comments left by more than 1 million users on CNN articles in 2012. The scientists couldn't gauge the mood of each of those people, but they looked at time stamps to make an educated guess—people tend to be in worse moods later at night and later in the week, and down-votes and flags on the CNN comments (indicating that people took issue with what was being said) followed this trend.
The researchers took it one step further and examined the context in which each flagged comment was being made. They found that a commenter was more likely to produce a flagged post if another one of their comments had been recently flagged or if they had participated in another online discussion in which there were other flagged comments.
It’s a spiral of negativity,” said Stanford's Jure Leskovec, an author of the study, to the university. “Just one person waking up cranky can create a spark and, because of discussion context and voting, these sparks can spiral out into cascades of bad behavior. Bad conversations lead to bad conversations. People who get down-voted come back more, comment more and comment even worse.”
When the researchers set up an algorithm to predict whether comment authors would create troll posts, the strongest predictor was whether the previous comment in the discussion had been flagged. Information about the commenter—their history of flagged posts and the time of day they were posting—was much less of a predictor. That means that no matter who we are, we're more likely to be trolls if other people are trolling around us.
How can we stop it?
A lot of news sites are combating hostile comment sections by simply removing the option to comment on articles. The researchers suggest media companies rethink that strategy and instead make rules that align with the results of burgeoning research on online discourse. For example, "shadow banning," hiding a troll's comment from other users without letting the troll know. That spares other commenters from a hostile environment that could cause them to start getting nasty, and prevents the offending troll from going on a commenting tirade.
“At the end of the day, what this research is really suggesting is that it’s us who are causing these breakdowns in discussion (online),” said Stanford's Michael Bernstein, another author of the study, to the university. “A lot of news sites have removed their comments systems because they think it’s counter to actual debate and discussion. Understanding our own best and worst selves here is key to bringing those back.”