Nathan at FlowingData points to an application on Facebook that aims to measures happiness. The application seeks to calculate the “Gross National Happiness” (GNH) of the United States by analyzing the ratio of positive to negative words found in users’ status updates on a daily basis.
It is an interesting project and certainly relates to the broader issue of using the data produced by social media technologies to measure and visualize emotions and feelings broadly. However, in its current form, the application falls well short of providing a true measure of gross happiness.
For example, I find it hard to believe that negativity fell generally during Q3 and Q4 of 2008 given the economic turmoil of the moment (see graph below). I suspect that, in general, users of Facebook will provide positive or neutral updates to their status, unless there is a singular, negative event that illicits strong emotion. This notion is illustrated by the fact that the highest ‘negative’ day was January 22, 2008–the day that Heath Ledger died and many people updated their status to reflect their sadness at his passing.
In the case of the economy, it’s highly unlikely that people would consistently harp on their negative feelings day in and day out. This should lead to an over-reporting of happiness and positivity. In fact, as some commentators on Nathan’s site suggest, we see massive spikes in ‘positivity’ during holidays where people general send good wishes via their status updates to all. Does that mean that as a nation we are that much happier? Perhaps, but the lack of downward fluctuations makes this data less compelling. In fact, the baseline GNH is pretty static, where the only real variation is in the positive direction. This isn’t so much a measure of “gross national happiness” as a time-series on the use of positive words in status updates.
Still, this project is an interesting first step in terms of coding the social web and using user-generated content to draw meaningful conclusions about perception, emotion, and feeling.