Twitter Study Reveals the Good Times and the Bad…By the Hour

New research is using Twitter to make maps of the national mood, according to The New York Times. The project, “Pulse of the Nation: U.S. Mood Throughout the Day Inferred from Twitter,” uses these social media messages to create maps that show mood levels across the United States over time.
The data — collected by researchers at Northeastern University and Harvard University — revealed interesting, if predictable, information about the national mood based on tweets. Americans are happiest on the weekends, with peak cheerfulness on Sunday mornings. Spirits take a dip on Monday and then remain lower for the next four days. Thursday evenings received the lowest mood rating, but scores moved steadily up throughout Friday as the work week waned.
Within each day, the research supported the idea of the afternoon slump: Twitter users recorded the happiest tweets in the early morning (6 a.m.) and late evening (9 p.m.). West coast mood patterns were usually three hours behind east coast patterns, in line with the time difference.
To compile their mood maps, the researchers used 300 million tweets taken from September 2006 to August 2009. The mood of each tweet was determined according to Affective Norms for English Words (ANEW), a program provided by the NIMH Center for the Study of Emotion and Attention. The number of positive or negative words each tweet had determined the overall mood score of that message.
Of course, the project can’t quite track the national mood since not everybody uses Twitter. The ANEW process also probably can’t detect sarcasm all that well. And since the most recent data used is almost a year old, it’s possible to think that maybe Twitter users in the United States have gotten happier — or at least less predictable — in the past year.