Things can get nasty on Twitter. Emotions run high and, by definition, they're there for the world to see. Most of us can only take so many intense Tweets before we decide that real-life isn't so bad after all.
For Rui Fan, however, there just aren't enough emotion-soaked Tweets in the day. He works in the State Key Laboratory of Software Development Environment of Beihang University. His specialism - the analysis of emotion as it's played out on social media.
In his latest analysis, Rui and his team identified over 109,000 Tweets that contained an 'I feel...' statement. Some of these were positive (I feel awesome) and some were negative (I feel terrible). Once he'd identified these Tweets, he was able to analyse the emotional content of the Tweets written in the 6 hours before it and 6 hours and after it. This meant he could track how quickly the positive or negative emotion had built up and how fast it then calmed down.
How do you analyse the emotional content of Tweets?
When you're working in the State Key Laboratory of Software Development Environment this is no problem. You simply use an established algorithm to check for about 7,000 emotion words previously rated for their positivity or negativity.
Naming negative emotions works more quickly
Rui and his team found that positive emotions were fairly symmetrical around the 'I feel...' Tweet. In other words, it took about the same time to build up to the 'I feel...' Tweet as it did to calm down after it. Notice here that naming the emotion was the turning point - after people came out and stated how they were feeling the emotion level began to decline.
Negative emotions were a little different: the build up was slower but the calming down after the 'I feel...' Tweet was steeper than for positive emotions. So naming negative emotions seems to work more quickly in diffusing an emotional episode than naming a positive emotion.
The emotion-naming effect is more powerful for women
Especially for negative emotions, there was some evidence that the decline in emotion intensity after the 'I feel...' Tweet was steeper for women than it was for men.
The pros and cons of big data analysis
The conclusion here is that naming emotions works to regulate how we feel and is especially useful for diffusing negative emotional experiences. And this is from a sample of 74,487 Twitter users.
However, the researchers themselves point out a few problems with their research. Not least is the problem of knowing the degree to which emotions played out on a social media stage are real or ...staged.
But still, this big data, together with the growing body of evidence, suggests that naming emotions is an effective way of regulating them. It's a simple, effective, evidence-based Psych Hack.