Over 40 million Facebook status updates are posted across the world every day. Status updates are simply short messages you post on your Facebook website prompted by a random question such as “What are you doing today?” “What’s going on?” or “How are you feeling?” Your Facebook friends get to read your status updates. Facebook makes available collections of such messages (anonymised) for researchers to analyse, providing a rich database of short texts indicating trends, obsessions and sentiments.
One use of this data is to infer the mood of a nation or community. Twitter and other microbloging sites provide similar databases. As reported in the New Scientist, researchers have mined twitter feeds to purportedly capture the mood of people across whole continents.
How happy are we? Researchers have devised algorithms that identify the frequency of occurrence of positive and negative words and phrases to derive indices on a happiness scale. The process focusses on very large numbers of short text messages rather than what individuals are feeling.
Perhaps social mood is about aggregation in any case, and averages — defining mood as an average feeling within a group of people. It comes as no surprise that there are sharp spikes in positive mood as bloggers wish one another Happy New Year and Merry Christmas. “Happy” and “Merry” are words, but their presence indicates a culturally embedded propensity for positive feelings between human beings.
It seems that we human beings are inclined to be be positive and happy anyway. According to some researchers, human beings exhibit a bias towards positive emotions. In most cultures, the instinctive response to “How are you?” is “I’m well thanks.”
If you imagine a scale between “sad” and “happy,” with “happy” at the upper end, then most people will place themselves somewhere above the mid point, ie above the neutral state. There’s a lot here about social convention and group solidarity, but psychologist Ed Diener states that most people report feeling slightly happier than the mid point — and really do.
Why are you so happy?
Why do we have this positive bias? Diener proposes that amongst other benefits, this bias actually sharpens our awareness to dangerous and unpleasant circumstances, and improves our chances of survival. It allows negative events to stand out even more strongly. If we incline to a positive mood, then against that background we are more likely to recognize and respond to danger, and the related emotions of fear and anxiety when they occur. So we are inclined to feel good so that we can really feel bad.
Our positive disposition apparently also makes us inclined to explore. Insofar as positive moods enhance our tendency to take risks and approach the unknown, then they help maintain our curiosity: the tendency to broaden our experience.
According to Diener, all of human history is caught up in this positive drive: “Human approach tendencies are manifest in the rapid exploration and settlement of new frontiers and in the unremitting invention of new ideas and institutions throughout human history.” This positive inclination expands from a local survival mechanism to embrace the whole of human history.
There’s the obvious question of what motivates such study, and how the results might be used. In benign political contexts the results could help in directing state resources to areas where people are unhappy. Or such statistics could be used to confirm the return on expenditure. One such study maps the happiness index derived from Twitter analysis in the various boroughs across London, claiming that the results correlate with well-being maps derived from official census data. Perhaps governments can gather well-being data without asking people to fill in a form at census time.
Such techniques don’t capture all there is to know about happiness. With far less reliance on masses of empirical data, Sigmund Freud suggested that we don’t only do things for the pleasure they bring. There’s deferred pleasure. We are also prone to compulsions that we have trouble describing as pleasures, such as repeating unproductive actions over and over — obsessive hand washing, pacing up and down, shopping, and updating Facebook statuses.
References
- Bradley, Margaret M., and Peter J. Lang. 1999. Affective Norms for English Words (ANEW): Instruction Manual and Affective Ratings (Technical Report C-1). Gainesville, FL: The Center for Research in Psychophysiology, The University of Florida. http://www.uvm.edu/~pdodds/files/papers/others/1999/bradley1999a.pdf.
- Diener, Ed, and Carol Diener. 1996. Most people are happy. Psychological Science, (7) 3, 181-185.
- Freud, Sigmund. 1990. Beyond the pleasure principle. In A. Richards (ed.), The Penguin Freud Library, Volume 11: On Metapsychology: 269-338. Harmondsworth: Middlesex: Penguin.
- Kramer, Adam D. I. 2010. An unobtrusive behavioral model of “Gross National Happiness”. Proc. ACM, Computer Human Interaction, Atlanta Georgia: 287-290.
- Peale, Norman Vincent. 1953. The Power of Positive Thinking. Kingswood, Tadworth, Surrey, UK: Cedar.
- Quercia, Daniele, Jonathan Ellis, Licia Capra, and Jon Crowcroft. 2012. Tracking “Gross Community Happiness” from Tweets. Proc. ACM, Computer Supported Collaborative Working, Seattle, WA: 965-968.
Notes
- Satisfaction With Life surveys require respondents to give number scores to such propositions as: In most ways my life is close to my ideal. The conditions of my life are excellent. I am satisfied with my life. So far I have gotten the important things I want in life. If I could live my life over, I would change almost nothing.
- It’s disputable whether or not the words people use are accurate indicators of a state of mind, but words do seem to matter. Hence the pop psychology advice: “While dressing or shaving or getting breakfast, say aloud a few such remarks as the following: ‘I believe this is going to be a wonderful day. I believe I can successfully handle all problems that will arise today. I feel good physically, mentally, emotionally. It is wonderful to be alive …’” (Peale, p.83).
- An interesting 1999 report has a table of words and their positive or negative valence (ie their association with positive emotions) as rated by human subjects. Words like “justice” get a score of around 8 out of 10, without much variation; “nasty” scores about 3.5 with a fair bit of variation. “Building” rates lower than “car” and “art,” and is on a par with “body,” but is above “bus.” “Architecture” is not there, nor are swear words listed.
- Also see other posts tagged with mood.
Reflections on Technology, Media & Culture 
I wonder if in this way a general ‘mood’ of i.e. a specific period of time can be tracked. i.e. of a specific decade. And I wonder how this general mood is related to the popular culture of that time. Like in music they say that generally that decade was more upbeat than this decade, and so on. So is this ‘mood’ (communicated and tracked in social networks) related to the mood of the popular culture of the time?
I also wonder about the interaction that posts of different moods initiate. For example do too ‘happy’ and too ‘sad’ posts generate more immediate responses than a ‘neutral’ (or close to neutral) post?
Hi Anastasia. What I’ve read about the general inference of mood from social media data is pretty general and broad brush. Mood is difficult to pin down anyway, which makes it kind of interesting, and a fertile area for speculation and exploration. As well as happiness, it would be interesting to explore melancholia through these media, a literary and productive kind of mood of sombre reflection — perhaps identifiable in London boroughs with a high proportion of poets, or bankers. I guess your last point about response times would involve an analysis of threads. I don’t know that status updates require a response, but some people use Facebook as a discussion forum. No doubt there are researchers working on all that data.
For a lot of people who may, for example, have work colleagues as part of their circle of “friends” on Facebook, status updates are highly self-censored and staged utterances. Personally, I would never say anything about my personal life on Facebook, and it’s pretty unlikely that I would write a negative utterance, given who’s reading, the unpredictability, and the tiny amount of control I have over it. I doubt I’m alone. It seems like a huge stretch, and faulty logic, that some researchers are trying to extrapolate people’s “mood” based on their Facebook utterances. Those just show how people feel they ought to present themselves in public — not how they actually feel.
Interesting article, but I agree with svenddottir.
I think people tend to post positive messages in order to present themselves as happy, successful, creative etc. I know friends who disappear from Facebook when they are feeling depressed – both because the sight of so many ‘happy’ people makes them feel worse, and also because they don’t have the capacity at the time to live up to this level of expectation.
It’s an interesting situation when we feel compelled to share our status with the world, yet most of us are able to take moment to honestly ask ourselves “how am I?”
Jung also defined the ‘shadow aspect’ which, being an unconscious aspect of the personality, tends to contain the traits we are least willing to embrace. This split between what we are unwilling to accept and what we think we should embody, can throw up a lot of problems. I wonder what role social media is playing here?
Excellent and sober comments from Svenddottir and Sarah. Insofar as there’s any mood mapping going on in the examples given in the article, it could be that these are “meta-moods,” ie aggregations of how people feel about describing, representing, faking, joking about, and being sarcastic about some mood or other at a particular time and place. In other words, mood is much more subtle than what’s implied by someone saying they feel great on a publicly readable web page. It looks like we can abandon the idea that social media provide windows into the psyche of a community, or people’s private lives and feelings. I had a shot at trying to understand meta-moods here: https://richardcoyne.com/2012/08/04/melancholy-and-media/, but not collective meta-moods.
Note the (un)happy planet index published online by the NEF (the new economics foundation). There’s a test you can do. It’s really about “sustainable well being.” http://www.happyplanetindex.org