Scientists Chart Worldwide Moods Using Twitter
How does the whole world feel right now? Just ask Twitter. While moods may feel like discreet and personal experiences, researchers at Cornell University propose that they fluctuate in predictable patterns on a global scale. And they are using the Twitter database to prove it.
In a study published today in the journal Science, sociologists Michael Macy and Scott Golder collected about half a billion messages posted by over 2 million English speaking Twitter users across the world, then treated the data as a huge bag of words. Using a homespun computer program, they sorted the words according to positive or negative categories and used them to rate the mood of the Twitter population .
Using their technique, Macy and Golder can take an informed look at how people are feeling in 84 countries at any time of the day.
"When you look at the overall pattern, what you see is a morning peak," Macy told InnovationNewsDaily.
Regardless of location, people seem to wake up in a relatively good mood. As the day wears on, the Twitter messages feed out a more negative score with a final lift again at the end of the day.
At first, the researchers assumed that the grind of the workday was bringing everyone down. But then, they noticed the same ups and downs over the weekend when most people are off.
"It's clearly not work that produces this pattern," said Macy.
He noticed two important differences between the weekend and the weekday. People seemed to wake up later on Saturdays and Sundays. And while their moods dipped toward the middle of the day, the baseline was higher, meaning they started at a higher place and never reached the lows that they did during the week.
It's possible, said Macy, that getting up early to drag themselves to work is what makes people most cranky. On the weekends, however, people have the luxury to follow their natural sleep cycles and this could explain why they stay in a better mood overall.
In the long run, using large, open-source social networks like Twitter as global databases could provide solid insight on how a society changes the behavior of individuals. But there is probably not enough evidence to start fiddling with work hours just yet.
For now, it's at least safe to feel a bit of solidarity at noon when angst and ennui settle in across the lunchroom.
"Maybe it's comforting to know that this is quite natural and that millions of people around the world are feeling the same thing," Macy said.