Facebook is a friendly distraction for most of us, but that doesn't mean it can't also be a tool: Facebook's been exploring those billions of status updates as a social data source for example. The data on happiness indices around the world are fascinating.

Back in October 2009 Facebook began this experiment, using anonymized status updates from its U.S. users and correlating key words inside them with the Gross National Happiness Index (a movement dedicated to assessing how cheery or upset a nation's citizens are.) Essentially Facebook's data team picked key phrases that relate to positive or negative emotions, made a frequency count of their occurrence and look at how the resulting data trends over time.

Some of the data isn't all that surprising, but it does prove the validity of the experiment: For example the "happiness" count tends to go up towards the end of each week and peak at the weekend. Nobody, but nobody, prefers workdays and Mondays. Facebook had to tweak the key words it searched for due to different populations in each of the new nations it's been analyzing (the U.K., Canada, and Australia) and to obvious "cultural differences in how people use language," as their blog posting puts it. It would be unusual to see Brits saying they had a "bonza" morning, just as for a Canadian to note their afternoon was "a little sub-par."

But the headline conclusions are actually amusing:

• Canadian's happiness about Thanksgiving peaks the evening before Thanksgiving day. Since the day itself is a Monday, it seems the Canadians are subject to the same Monday woes as the rest of us
• Canadian happiness is affected by a little cross-border feeling from the U.S.--happiness indices show small peaks on July 4th and U.S. Thanksgiving
• The U.K.'s happiness, while still tracing up and down, is the most smooth in terms of happy peaks at holiday days. Facebook's analysis suggests it's due to the "heterogeneity of U.K. bank holidays" (jealous of that much, Facebook analysts?) but I suspect its that good old British reserve showing in the data
• The Aussies have a huge drop in happiness that correlates to so many users Facebooking the word "sorry" last year when their Prime Minister issued a Parliamentary apology to the terrible treatment throughout modern Australian history of its indigenous population
• While everybody likes Valentine's day, it looks like the Canadians like it the most (are they known for being big romantics?) And while Canada also loves Halloween, that festival day is undetectable in the Australian graph, and only shows a slight bump in the U.K. chart (where the thing originated, of course, as a much more sober Celtic affair, with less emphasis on candy and pranks.)
• The Christmas holiday stands out too, but since Facebook's separate and different analysis of the data between nations precludes international comparisons, we can't really say if this reflects differing national interest in religious events. The fact that Hanukkah and Diwali are at roughly the same time will also play into the stats of course

Potentially the most useful piece of data from Facebook's study is that negativity is decreasing everywhere. While the blog posting notes that this could be due to changing demographics (as more older people are using the service now) it could also be taken as an upswing indicator: The end of the economic gloomy patch, perhaps. Money can buy you happiness...

And thus is revealed the true value of this analysis. Facebook's demonstrated that the vast historic record of status updates is a potential goldmine of information that could easily be raked through by sociology analysts keen to work out when it's best to deliver an advert for particular products, or perhaps even to promote a particular political message.

[Facebook]

To keep up with this and other news, happy, sad or just a bit "meh" in nature, follow me, Kit Eaton, on Twitter. That QR code on the left will take you to my Twitter feed too. I am generally a cheery Tweeter :)