sad face

A new study by researchers at the University of Vermont uses tweets to determine the level of happiness among people all over the world.

The Twitter hive may hold a clue to the general mood of the global mind. According to a new study in the PloS ONE journal, tweets have been steadily trending towards unhappiness over the past three years.

Happiness is important, for some it’s the purpose of society itself. However, this important societal metric is hard to tie down, as many traditional means of measurement have a hard time digging out honesty. Indicators like gross domestic product are far more easy to quantify.

A group of scientists from the University of Vermont say they have figured out how to look over the “collective shoulder of society” and measure happiness in a natural way, by using Twitter to aggregate expressions and track people’s mood patterns in the micro-blogging wild.

Peter Dodds and his team say that happiness is going down. The team writes in the journal, “After a gradual upward trend that ran from January to April, 2009, the overall time series has shown a gradual downward trend, accelerating somewhat over the first half of 2011.”

swear words graph via plos one

The data is collected from over 46 billion words in tweets form 63 million twitter users worldwide. From “pancakes” to “suicide” the researchers then create ratings for words using Amazon’s Mechanical Turk. From a scale of 1 to 9, users rated ten thousand of the most common English words on how much happiness a word evokes. The study notes, for example, that “laughter” was scored at 8.5, while “greed” scored a 3.6 and “terrorist” was given a 1.30.

In the short term, Dodds’ team notes that the happiness signals peak over the weekend, and ebb at the beginning of the week. Happiness also flutters throughout the day, from early morning to nighttime as the mind unravels. Happiness fluctuates over the year, jumping up at holidays and dropping during shocks to routine such as Patrick Swayze’s death or the Japan tsunami. However, Dodds insists that, overall there is a drop in happiness.

The happiness sensor does have its flaws as Dodds agrees that this is Twitter-centric. He says,“it does skew toward younger people and people with smartphones and so on—but Twitter is nearly universal now. Every demographic is represented.”

Via University of Vermont

Showing 1 comment

  1. sdey001 at 8:34pm 21st December 2011 The rate with which ‘happiness’ is being reported as “going down” is an interesting observation. I seem to have rather profound difficulty to define happiness and the research method that has been applied to project the magnitude of happiness as a dependent variable. All that said, I still like the tone and style of the ‘social study’ which may lead to other scholarly investigations. However, there is a prominent issue that concerns me. Using data mining techniques on large scale text data may require the ‘source data’ records. Other elements in that record probably needed to be validated to apply appropriate research model. That brings my million $ question! Are all data in Twitter or any Social Media open for analysis and thus may not be subject to client privacy? Is it not like e-mail privacy breach? The researcher claims “it does skew toward younger people and people with smartphones and so on—but Twitter is nearly universal now. Every demographic is represented.” So, do I know every demographics of each record? I am not sure and I may be far off from even raising this as anything to worry about. Being involved in healthcare related studies and systems for over 25 years, that caught my eyes. My apologies! Dr. Dey
Close Suggestion Google+ brand pages start appearing in Google search results
View Article