Study: Everyone Is Guilty Of Trolling On The Internet
Acting like a “troll,” or a person who deliberately posts provocative or offensive content, on the internet is actually quite normal, according to a new study.
The study, which is partially titled “Anyone Can Become a Troll,” revealed that while the most intense violators of community norms are a serious problem, sensible people often write comments that someone else may consider inappropriate or overly aggressive.
“We adopt a definition of trolling that includes flaming, griefing, swearing, or personal attacks, including behavior outside the acceptable bounds defined by several community guidelines for discussion forums,” the study reads.
They argue that administrators of websites and platforms should not just focus on the repeat offenders in order to thwart trolling.
“It’s tempting to believe that all the problems online are due to someone else, some really sociopathic person,” Michael Bernstein, one of the four collaborators of the research, told The Wall Street Journal. “Actually, we all have to own up to this.”
The computer scientists at Cornell and Stanford universities who conducted the study say that not all trolling can be blamed on just a few bad eggs, and attribute the relative pervasiveness to online platform’s “deindividualization” and reduction in accountability.
“This disinhibition effect suggests that people, in online settings, can be more easily influenced to act antisocially. Thus, rather than assume that only trolls engage in trolling behavior, we ask … Can situational factors trigger trolling behavior?” the study continues.
The researchers set up an experiment where participants would complete a quiz and then engage in an online discussion. The test consisted of 15 open-ended questions, including logic, math, and word problems.
“The goal of the quiz was to see if participants’ mood prior to participating in a discussion had an effect on subsequent trolling,” the study continues.
The study also helped to corroborate its findings with an analysis of over 16 million posts on “CNN.com,” which showed “one out of four posts flagged for abuse was authored by users with no prior record of such posts.”
“We find that users who had a post flagged in a prior discussion were twice as likely to troll in their next post in a different discussion,” the study continues.
Furthermore, researchers say that future work should distinguish between two types, both people who constantly troll and people who troll situationally.
For the latter, trolling could stem from personal mood, which can be affected by a number of factors like “unrelated aversive events,” like a relationship gone bad or workplace frustration.
“Circumstances that influence mood may also modify the rate of trolling. For instance, mood changes with the time of day or day of week. As negative mood rises at the start of the week, and late at night, trolling may vary similarly,” the study continues, explaining why the majority of trolling occurs really late at night, and on Monday’s.
One news site in Norway is forcing readers to complete a quiz on the content of a story before writing a comment, in an attempt to limit the toxicity of conversations and foster constructiveness.
Facebook has somewhat stringent rules to curtail trolling and purge its platform of provocative content, but is sometimes accused of going too far, including an instance in which a company’s lingerie ads featuring a transgender and amputee model were removed. (RELATED: Cancer Awareness Group Draws Square Breasts For Video After Facebook Censors It)
Google parent company Alphabet recently announced that it was working with a number of media outlets, like The New York Times, The Guardian, and The Economist, in order to develop machine learning technology that will cultivate more productive online conversations by automatically identifying “toxic comments.” (RELATED: Twitter Says It’s Developing New Tools For Online Censorship)
Twitter has been under pressure to censor users for posting disrespectful or downright trollish tweets. Contrarily, free speech advocates argue that the social media company shouldn’t remove “trolls” because their acts are usually protected under the First Amendment. (RELATED: Twitter Doesn’t Rule Out Banning Trump If He Gets Too Unruly)
While a number of websites have explicitly forbidden particular users from commenting, or completely removed the feature altogether, researchers disagree with that tactic.
“Considering measures that mitigate the situational factors that lead to trolling may better reflect the reality of how trolling occurs,” the study concludes.
Send tips to firstname.lastname@example.org.