Twitter study points to global mood swings
Twitter is many things to many people — a social place, a place of business, a media platform — but now it has a new role as a mood meter, with researchers finding most people worldwide have the same mood rhythms.
Cornell University researchers used a text-analysis program called the Linguistic Inquiry and Word Count to analyze 509 million tweets and “gauge the positive affect (enthusiasm, delight, activeness, alertness, etc.) or negative affect (distress, fear, anger, guilt, disgust, etc.) expressed by the text,” according to a news release about the study.
The researchers found “people tend to be more positive on weekends and early in the morning. In general, individuals awaken in a good mood that slowly deteriorates as the day progresses, which is consistent with the effects of sleep and circadian rhythms.”
“On weekends, those early-morning good moods are delayed for two hours, suggesting that people sleep later on those days.”
Twitter moods and the rhythm of those moods are about the same in among the 2.4 million users in 84 different countries surveyed over a two-year period.
Scott Golder, Cornell graduate student in sociology, and Michael Macy, Cornell professor of sociology, conducted the study, which will appear in the Sept. 29 edition of Science that is published by AAAS, a non-profit science society. Golder said there is a lot going on when it comes to the timing of the moods.
“Many lay people who I talked about this said, ‘Of course, people go to work and they don’t have very high positive affect at work,’” said Golder.
“It turns out that this isn’t the whole story. We saw that the shape of the rhythm was exactly the same on Saturday and Sunday when many people are not at work. So clearly something else is playing a role here, whether it’s sleep or biological rhythms…” he said.
Studying people via Twitter allows researchers the chance to learn about the habits of a large group of people acting in their own environment without disturbing them.
There are limitations to using Twitter, said the researchers, who noted “that unlike laboratory studies, we have little data on conditions that are known to influence mood, including demographic and occupational backgrounds that may influence when and how much people sleep, the level and timing of environmental stress, susceptibility to affective contagion, and access to social support.”
