Happiness Is on the Decline, at Least According to What We Tweet


Scientists from the University of Vermont have undertaken an in-depth, extremely thorough examination of the emotions of the Twitterverse, and they’ve found that happiness is…decreasing. In the journal PLoS ONE, they write that “a gradual downward trend” is evident over the first half of 2011, following a gradual upward trend in 2009. What does this mean? “It appears that happiness is going down,” said Peter Dodds, lead author on the study. Ouch. That’s a bummer. However, the researchers’ methodology and learnings are fascinating.

Dodd and his colleagues looked at more than 46 billion words contained in the tweets of 63 million Twitter users. (Recall the Hack Day project we wrote of recently that determined which subway stops were the “happiest”? This is somewhat similar.) The words collected ranged wildly, of course, given what different people tweet. The scientists had volunteers rate the sense of happiness of the various words using Amazon’s Mechanical Turk. “Laughter” got a high happiness score; “terrorist” a low one — of course, “laughter” could be an unhappy word if someone’s laughing at you, and Osama bin Laden’s death might be happiness-producing, which is to say…there’s a measure of human error given lack of context here.

After the test words were ranked, the scores were applied to all the words collected from Twitter. Those words were also categorized by date and time and geography, where applicable, to show how happy a New Yorker was on a Saturday night, say, as opposed to a Monday morning.

In general: “Stretched out over the last three years, these patterns of word use show a drop in average happiness.” People were happiest over weekends and least so early in the week. Interestingly, they were happier in the morning than late at night, though there was another happiness spike at 7 or 8 at night (after that happiness levels plummet, and midday, things are pretty dreary).

Keep in mind, of course, that this doesn’t account for people who are made happy by bitching about things online. No one can account for them, really.

Many interesting charts about all this here.

Temporal Patterns of Happiness and Information in a Global Social Network: Hedonometrics and Twitter [PLoS ONE via Newswise]