Students’ tweets could reveal more about their happiness than they might realize.
Researchers at the University of Vermont are attempting to quantify America’s happiness from an unconventional source of data — tweets.
The results of a study released last week titled, “The Geography of Happiness” showed Michigan’s Twitter users are particularly unhappy, ranking 45th out of 51 states and Washington, D.C.
The study examined more than 10 million of Twitter’s publicly available, geotagged tweets.
The researchers analyzed the word content of tweets, assigning each word a value of happiness based upon the Language Assessment by Mechanical Turk word list.
Gauging Tweet Happiness
(Scale of 1 to 9, with 1 being saddest and 9 being happiest)
-Lansing’s average: 5.96
-Ann Arbor’s average: 5.91
-All cities : 5.99
Most positive words in Lansing:
Most negative words in Lansing:
SOURCE: UNIVERSITY OF VERMONT
James Madison College freshman and Maryland resident Joe Mack speculated why Michigan placed low in the happiness ranking.
“It’s freaking cold,” he said, later adding “the economy doesn’t seem too hot comparatively.”
Lewis Mitchell, one of the study’s authors, said excessive use of profanity made Michiganians rank comparatively sad. They tweeted subtly negative words such as “hurt,” “don’t,” “battle” and “falling” more frequently and “thanks” and “awesome” less often.
“One of the major facets of this study was to take this measure of happiness and to correlate it with existing characteristics of cities,” Mitchell said. “We tried to relate this to obesity. From there we found that as obesity goes up in a city, happiness goes down.”
According to U.S. News and World Report, Michigan is the 5th most obese state in the U.S.
The study also drew connections between words and socioeconomic factors. For example, the word “cafe” correlated strongly in populations with both low obesity rates and high percentages of bachelor’s degrees.
“The words which correlate negatively with education are generally shorter, with no words longer than two syllables appearing in the list,” the study said. “The more technical terms appearing in (areas with higher education) are more employment-oriented and suggest more complex and abstract intellectual themes.”
Words associated with low levels of education revolved around interpersonal relationships, with the word “me” bearing the strongest correlation.
Mitchell hopes aggregated data from Twitter can be used to provide a “real-time indicator of how a city or state is faring.”
Kinesiology freshman Michelle Sondgerath viewed her Twitter as an anthology of her friends’ lives, presenting a mixed bag of both positive and negative.
“They tweet about stuff they’re doing, stuff people say, stuff that’s going on in their lives — some good, some bad,” she said.