"Long before it's in the papers"
September 29, 2011

RETURN TO THE WORLD SCIENCE HOME PAGE

Scoop from Twitter data: People are happiest in the AM, sour on work

Sept. 29, 2011
Courtesy of Cornell University
and World Science staff

Us­ing Twit­ter to mon­i­tor mil­lions of peo­ple’s at­ti­tudes, re­search­ers have found that hu­mans all over the world tend to wake up in a good mood, but feel grouch­i­er as the day wears on.

Work hours, per­haps un­sur­pris­ing­ly, seem to be the times most widely as­so­ci­at­ed with bad mood, the new study sug­gested.

Re­search­ers an­a­lyzed the text mes­sages gen­er­at­ed by the free so­cial mes­sag­ing tool, which lets peo­ple stay dig­it­ally con­nect­ed through up­dates called “tweets” up to 140 char­ac­ters long. The mes­sages are trans­mit­ted through cell phones, the Web or e­mail.

The stu­dy, by re­search­ers at Cor­nell Uni­vers­ity in New York, an­a­lyzed tweets from 2.4 mil­lion peo­ple in 84 coun­tries. The re­sults ap­pear in the Sept. 29 is­sue of the re­search jour­nal Sci­ence. By track­ing tweets over two years, the in­ves­ti­ga­tors found that work, sleep and the amount of day­light play a role in shap­ing cy­cli­cal emo­tions such as en­thu­si­asm, de­light, alert­ness, dis­tress, fear and an­ger.

Sci­en­tists have long known about such rhythms, but pre­cise and real-time read­ings were un­ob­tain­able be­fore the rise of so­cial me­dia, ac­cord­ing to the re­search­ers, so­ci­ol­o­gy grad­u­ate stu­dent Scott Golder and so­ci­ol­o­gist Mi­chael Macy.

Us­ing Twit­ter in con­junc­tion with lan­guage mon­i­toring soft­ware, they found two daily peaks in which tweets rep­re­sented a pos­i­tive at­ti­tude – rel­a­tively early in the morn­ing and again near mid­night, sug­gest­ing mood may be shaped by work-related stress. Pos­i­tive tweets were al­so more plen­ti­ful on Sat­ur­days and Sun­days, with the morn­ing peaks oc­cur­ring about two hours lat­er. This im­plies peo­ple awak­en lat­er on week­ends, said the re­search­ers.

These pat­terns were re­flected in cul­tures and coun­tries through­out the world, but shifted with the dif­fer­ence in time and work sched­ule. For ex­am­ple, pos­i­tive tweets and late-morn­ing mood peaks were more prom­i­nent on Fri­days and Sat­ur­days in the Un­ited Ar­ab Emi­rates, where the tra­di­tion­al work­week is Sun­day through Thurs­day, ac­cord­ing to the pa­per.

Golder and Macy al­so tracked glob­al at­ti­tude on a sea­son­al ba­sis to de­ter­mine if “win­ter blues” is re­flected in Twit­ter mes­sages. While no cor­rela­t­ion was disco­vered be­tween ab­so­lute day­light and mood, they found a cor­rela­t­ion when ex­am­in­ing rel­a­tive day­light, such as the grad­u­ally de­creas­ing day length be­tween the sum­mer and win­ter sol­stices.

* * *

Using Twitter to monitor millions of people’s attitudes, researchers have found that humans all over the world tend to wake up in a good mood, but feel grouchier as the day wears on. Work hours, perhaps unsurprisingly, seem to be the times most widely associated with bad mood, the new study suggested. Researchers analyzed the text messages generated by the free social messaging tool, which lets people stay digitally connected through updates called “tweets” up to 140 characters long. The messages are transmitted through cell phones, the Web or email. The study, by researchers at Cornell University in New York, analyzed tweets from 2.4 million people in 84 countries. The results appear in the Sept. 29 issue of the research journal Science. By tracking tweets over two years, the investigators found that work, sleep and the amount of daylight play a role in shaping cyclical emotions such as enthusiasm, delight, alertness, distress, fear and anger. Scientists have long known about such rhythms, but precise and real-time readings were unobtainable before the rise of social media, according to the researchers, sociology graduate student Scott Golder and sociologist Michael Macy. Using Twitter in conjunction with language monitoring software, they found two daily peaks in which tweets represented a positive attitude – relatively early in the morning and again near midnight, suggesting mood may be shaped by work-related stress. Positive tweets were also more plentiful on Saturdays and Sundays, with the morning peaks occurring about two hours later. This implies people awaken later on weekends, said the researchers. These patterns were reflected in cultures and countries throughout the world, but shifted with the difference in time and work schedule. For example, positive tweets and late-morning mood peaks were more prominent on Fridays and Saturdays in the United Arab Emirates, where the traditional workweek is Sunday through Thursday, according to the paper. Golder and Macy also tracked global attitude on a seasonal basis to determine if “winter blues” is reflected in Twitter messages. While no correlation was discovered between absolute daylight and mood, they found a correlation when examining relative daylight, such as the gradually decreasing daylength between the summer and winter solstices.