Moods worldwide revealed by Twitter
An analysis of over half a billion "tweets" worldwide has confirmed that, regardless of country or culture, we're all in a better mood in the morning!
Researchers Scott Gelder and Michael Macy fed into a language-analysing computer programme the messages posted on the Twitter microblogging site by 2.4 million people from 84 different countries over a two year period.
Based on the words used, the "tweets" were divided up into those conveying either a positive or negative vibe, known as affect. Plotting the frequency of these good or bad vibrations against time of day, and then comparing the results geographically, the Cornell-based duo were able to create a two-year profile of how the daily moods of different cultures varied with time.
In general, the results reveal, we're all much happier early on in the day, with a significantly higher rate of positive tweets posted first thing. There's also a late flurry of positive sentiment just before bed. Predictably, weekends also tend to be associated with more a positive affect, although the peaks at weekends tend to occur later in the day, consistent with people having a lie-in. (Clearly the majority of weekend tweeters don't have young children). Interestingly, negative affect, which is lowest in the morning and grows across the day, is not linked to positive affect - the two sentiments are not at opposite ends of a single dimension but instead move consistently in parallel.
The longitudinal nature of the study also means that its perfect for studying another well-known mood-related problem: the "winter blues". By marrying up the trends in the data with the relative day lengths Gelder and Macy were able to test the theory that seasonal affective changes are not directly linked to the absolute length of a day per-se, but instead to the relative change in day-length and the timing of the onset of dawn, which is important for setting the body's internal clock. Absolute day-length, they found, makes little difference to mood, but shortening days are associated with people posting fewer positive comments while lengthening days produce more positively-loaded expressions.
Published in the journal Science this week, the sheer scale of this enterprise, and its cross-cultural and multi-demographic nature, overcomes many of the hurdles previously encountered by researchers trying to study how the moods of individuals varies across a day. So now you know when not to ring someone up with bad news...