The world is getting unhappier, according to Twitter
Some 46 billion words tweeted by 63 million users indicate that the world's happiness as measured through tweets is at its lowest since 2009.
'Tis the season to be jolly. And a lot of us are during the holidays, if statistical analyses of our tweets provide sufficient measure.
But overall happiness appears to be on a gradual decline since 2009, according to a University of Vermont analysis of some 46 billion words tweeted by 63 million users since 2009.
The team compared a wide range of words and phrases--including hahaha and lol--to "happiness scores" of the 10,000 most common English words. Words such as happy and laughter appear at the top of the 1-to-9 scale, while terrorist gets 1.30.
Not only did the team find a gradual downward trend of happy words since 2009, that trend actually accelerated during the first half of 2011 as well, with certain "negative" events (tsunami in Japan, swine flu outbreak) pushing those lows down even further.
"This is at the worldwide level, so the things that show up really have to be huge," Peter Dodds, lead author of the study, tells the media. "We're never universally accidentally happy, but we have repeated global shocks that generate a surge of negativity.... We're at our lowest point now in four years as far as our measure of happiness through Twitter goes."
It turns out that spikes in happiness as measured by the words we tweet consistently occur right before the weekends and holidays (think New Year's Eve and Fridays). Our anticipation of happy events, it seems, may make us even happier than the events themselves.
Dodds and his team say they plan to start a Web site in the next few months to let the general public mine the data themselves. For now, their study in PLoS ONE includes several graphs worth inspecting, some reminiscent of refrigerator poetry.
Of course, it's worth noting that by measuring tweets alone, the researchers aren't accounting for the emotions of all those who don't use Twitter. Still, the sheer volume of tweets makes the results worth at least some consideration.