This new study suggests sad tweets make you sad

Just as a hilarious photo or powerful quote might spread virally across Twitter, so too might emotion.

A new study from Indiana University found that emotions on Twitter appear to be contagious, with positive tweets inspiring users who read them to tweet more positively, and negative tweets begetting, you guessed it, more negative tweets. (The study is currently under review for the journal PLoS ONE.)

If this sounds like last year’s Facebook’s emotional manipulation fiasco all over again, fret not. Unlike the Facebook study, researchers didn’t exercise any control over what users saw in their Twitter feeds. Instead, using Twitter’s API, researchers pulled a random sample of 3,800 English-speaking Twitter users, then observed and analyzed how they appeared to respond to tweets that showed up in their stream. Using an algorithm that ranked how positive or negative the content of a tweet was, researchers found that on average those who tweeted negatively had been exposed to about 4.34 percent more negative tweets and those who tweeted positively had seen about 4.5 percent more positive content.

So if you want to bring out your dark side, start following @sosadtoday, but if you prefer to be more upbeat, go for @thehappyquote.

https://twitter.com/sosadtoday/status/613392435247480832

Past research has shown that a person’s happiness is significantly dependent on those they are connected to in real life. Emilio Ferrara, an Indiana University professor who studies network effects and headed up the research, was interested in recreating a more ethical version of the Facebook study to explore how emotional contagion translates to online social networks.

Ferrara said the correlation they discovered may be small, but it’s significant. It’s also a larger correlation than the Facebook study found.

If might seem obvious that reading something positive makes you feel happier, but that’s not what past researchers have found. Previous studies in the U.S. and in Germany found that all that positive, self-promotional content Facebook tends to feature at the top of your feed (“I’m getting married!” “I got a new job!” “I’m so happy and grateful for how awesome life is!”) tended to make users feel worse. German researchers called it “the self-promotion-envy spiral.” (Facebook’s emotion contagion study was motivated in part by hopes of debunking such research.)

Ferrara was interested in Twitter specifically because, while we may post less personal information on Twitter, recent neuroscience research (albeit research done by Twitter) has suggested we may actually be more emotionally-engaged on Twitter than elsewhere on the Web. In that study, researchers monitored the brain activity of 114 people while they surfed the Web, finding that browsing a Twitter timeline generated 64 percent more activity in the parts of the brain known to be active in emotion than other Web use.

Some tweeps are more emotionally vulnerable than others. Ferrara’s study found that about 20 percent of Twitter users were particularly susceptible to the power of Twitter’s emotional influence, with more than half of their tweets showing evidence of “contagion.” And just as many news organizations have discovered (spawning sites like Upworthy), people like spreading good news. The study found that positive tweets were much more contagious than negative ones.

So does this actually tell you how Twitter users feel? A major criticism of the Facebook study was that it inferred people’s moods based only on what they posted. Ferrara is well-aware that all they can really tell from what you tweet is, well, what you tweeted.

“Of course, it’s very difficult to understand whether a user’s tweet reflects their emotional status overall,” he said. So what’s the point of scientifically proving that a sad tweet might inspire you to tweet something sad?

One scenario, said Ferrara, is helping marketers more effectively reach social media users with their message. But Ferrara is more interested in how this kind of research might influence other kinds of messaging, like, say, how governments communicate during large scale disasters. Perhaps knowing what kinds of information comes across positively could help to disseminate emergency information effectively and quell panic. Social media conversations, he said, impact the physical world in many tangible ways.

“It’s really important because it has a very practical effect on every user,” said Ferrara. “Whenever you share something online that content doesn’t affect only you. In principle, it impacts everybody listening.”