Twitter Happiness in Decline
Uncovering and explaining temporal variations in happiness on Twitter.
This has been a pretty eventful year by recent standards, what with the protests and uprisings in the Arab world, natural disasters like the Japanese earthquake, and the economic gloom in the west casting an ugly shadow over it all.
But while this year may have been eventful, it certainly hasn’t done much for our level of happiness, according to a new study by the University of Vermont which links our moods with our tweets on the social networking platform, Twitter.
Researchers at the university believe that Twitter is able to track the general mood of the globe’s ‘connected’ population. And according to their results, our overall level of happiness has been steadily decreasing for the last three years.
Many people believe that the most important reason for living is to be happy and enjoy life to its fullest. After all, if everyone was unhappy and miserable all the time, what would be the point in us being here, really? But while the whole “meaning of life” debate will probably rage on until the end of time, there is no denying that happiness is important in our society. Yet it’s also one of the most difficult social metrics to measure, or at least it was until now.
However, the Vermont University scientists are convinced that they have stumbled upon a new method of measuring happiness accurately, simply by monitoring people’s ‘tweets’ on Twitter to track their overall mood. The study came up with some pretty interesting findings.
Lead researcher Peter Dodds says that unhappiness is on the rise. He said that the trend began back in January 2009, and has accelerated over the last, eventful twelve months.
Dodds and his team examined more than 46 billion words used in approximately 4.6 billion tweets in the study, which took place over 33 months from September 2008 till September 2011. To get an idea of the level of happiness of Twitter users, the team assigned specified words a “happiness score”, based on the belief that people use certain words according to their mood. For example, the word “laughter” was given a happiness score of 8.5, while words like “terrorist” (1.3) and “greed” (3.06) were given lower scores to reflect that they generally aren’t used by people who are in a good mood.
The team’s findings show that our overall level of happiness has experienced a steady decline since 2009, and this trend has accelerated over the last year.
Of course, our level of happiness does fluctuate – people tend to be much happier at the weekend, but this positive mood quickly evaporates once we get back to work on a Monday, reaching its lowest level by Tuesday. In addition, our happiness levels ebb and flow throughout the course of the day. The team tracked the use of obscenities (swear words) throughout the course of a day – these generally score quite low on the happiness charts. They found that swear word usage was quite low in the mornings (a sign perhaps that many people start the day with a more positive outlook), steadily increasing throughout the day and peaking by around midnight, which leads one to speculate that people tend to become more irritable and moody as the day progresses.
Our happiness is also affected by different events and periods throughout the year. More people tend display happiness during national holidays such as Christmas and Valentine’s Day, yet when bad news comes along this optimism is quickly deflated, like when the England football team was knocked out of the World Cup, the day Patrick Swayze died or when the Japanese earthquake struck.
Dodds is the first to admit that the study has some flaws however. He agrees that the results are skewed towards those who use Twitter – which is generally more well-to-do people who can afford to own smartphones and the like. Twitter is also used mainly by the younger generation, so the happiness levels probably reflect their mood more than that of the world’s senior citizens.
The findings do seem to contradict themselves as well, at least in some cases. It’s interesting to note that the unhappiest day recorded in the 33-month period was May 2, 2011, the same day that Osama Bin Laden was shot dead by US forces in Pakistan. Yet compare this with the scenes of jubilation that were splashed all over our TV screens, with hundreds of people dancing and celebrating the news outside the White House. This particular incident does call into question the “happiness score” system used by the team – the word “terrorist” was most likely used in a much happier context that day.
Still, overall the results are pretty convincing, and what’s more they are backed up by Facebook, who made their own attempt at measuring happiness. Many of the findings of Facebook’s Gross National Happiness Index seem to concur with the Twitter study – with happiness levels peaking at weekends and holidays such as Christmas and Valentine’s Day.