Twitter can help scientists peer into hearts worldwide to reveal how moods swing globally over time, a new study finds.
The microblogging service Twitter is often derided as a way for people to babble "tweets" about what they had for breakfast and other trivialities of daily life. However, in the sheer volume of messages now tweeted daily — an average of 230 million per day, according to September statistics — a growing number of researchers are now using Twitter to unearth insights into human behavior.
Twitter users cannot say much in a single tweet – there is a 140-character limit – but they update constantly, often using mobile devices. This means they often quickly report details about their lives — including disasters and other momentous events — in real-time. Unlike Facebook, tweets are also typically publicly available unless a user chooses to lock them. [Read also "20 Terrific Twitter Tips."]
A global mood ring
Sociologists Michael Macy and Scott Golder at Cornell University analyzed 509 million tweets from 2.4 million users in 84 different countries over a two-year period. They employed a text-analysis program known as the Linguistic Inquiry and Word Count for words with "positive affect," such as delight and enthusiasm, and negative ones, such as fear and anger.
The researchers found that people across the globe display similar rhythms to their moods despite very different cultures, religions and geographies. For instance, people tended to be more positive on weekends and early in the morning. Those who were analyzed that usually woke up in a good mood slowly deteriorated as the day progressed.
On weekends, these early-morning good moods were delayed for two hours, suggesting that people sleep later on those days. This was confirmed even in the United Arab Emirates, where people work Sunday through Thursday.
"People's moods are deteriorating not just on weekdays, but on weekends too, when most people aren't working, so this effect is not just due to work," Macy told TechNewsDaily.
Instead, these results might be consistent with the effects of sleep and circadian rhythms, he suggested.
"On weekends, people aren't woken up by alarm clocks for work, so they get to wake up naturally, which might help account for the elevated mood that we observed," Macy said.
In addition, the researchers tracked global attitudes on a seasonal basis to see if they could find signs of "winter blues." They found what might be a link between mood and when day length was gradually increasing or decreasing over time between the summer and winter solstices.
The scientists also investigated a group of "night owls," those most active between midnight and 6 a.m. These were slightly different in that their moods did not rebound in the evening.
The researchers did caution their study has limitations.
"Although the millions of Twitter users we had are a much more diverse sample to study than the small number of undergraduates in psychology labs often used for mood studies, we know that access to Twitter has age, income and education biases, so we're hesitant to generalize from these results," Macy said. "Having said that, the similarity of this pattern we see from India to Africa to Australia to the U.K. to Canada and the U.S., the similarity of results across diverse cultures, makes us confident that our results are robust."
Macy said his team is planning follow-up studies to better understand the reasons for the patterns they’re seeing.
"Are they shaped by sleep, by work, or by some combination?" Macy said. "We can analyze these messages to find clues about what days people are working and what time they go to work, and also about their demographic characteristics such as age, gender, and so on."
Macy and Golder detailed their findings in the Sept. 30 issue of the journal Science.