New research reveals the latest secret to happiness and low mood – how you use Twitter
Raj Persaud and Adrian Furnham
The contentment of those up to three links away from who you directly interacted with, could also impact on your happiness. In other words, positive mood appears to spread through a network and could even be contagious. Perhaps similarly, depression and other emotional problems, might be socially infectious.
Another key finding from the research was more connected users tend to be happier.
Catherine Bliss, Isabel Kloumann, Kameron Harris, Christopher Danforth and Peter Dodds
analysed nearly 40 million messages posted to Twitter. This revealed social network structure
and dynamics over 6 months.
The study entitled, ‘Twitter reciprocal reply networks exhibit assortativity with respect to happiness’, focused on interaction between people. The term ‘assortativity’ means that happiness is not random – those in higher spirits seem to find and link with each other – or interacting with the contented renders you more up-beat.
But just following someone was not the key focus of the study. Two people being ‘connected’ in some social sense occurred, according to this research, if both had replied to each other. This study focused on reciprocal interaction using Twitter. The authors believed they were studying twitter users who were part of a social network, similar to a set of connections in the physical environment, such as neighbours who you speak to in your neighbourhood, or colleagues at work.
Just published in the ‘Journal of Computational Science’, the authors describe ‘Twitter’ as an ‘online, interactive social media platform’ in which users post tweets, micro-blogs with a 140 character limit. Since its start in 2006, the authors explain, Twitter has grown to over 200 million accounts, (at the time the paper was submitted to the journal in October 2011) with some users having garnered over 10 million followers.
This study used a ‘hedonometer’ for measuring sentiment in text, which had previously developed and used before in similar research. 10,222 of the most used words in the English language, on a happiness scale from 1 to 9 (1 representing sad and 9 representing happy) were scored. The average happiness score of a word is the average from 50 independent evaluations. Examples are: ‘love’ as a word scores 8.42 on happiness, ‘special’ achieves 7.20, while ‘never’ drops to 3.34, ‘sad’ gets 2.38, and ‘die’ languishes at 1.74.
The study computed the happiness of each user by applying this ‘hedonometer’ to all tweets authored by the user. Each users’ collection of words reﬂects many messages – not just replies to those involved in reciprocal interaction.
So a tweet message such as ‘Vacation starts today yeahhhh’ – the word yeahhhh wasn’t coded, but ‘vacation’ is a word scored by many previous independent assessors in this kind of research, as more associated with greater ‘happiness’, than the other words in that sentence. As a result, ‘Vacation’ is allocated a happiness score of 7.92.
Using this way of measuring how happy users of twitter are, contentment levels are found to be more similar to their nearest neighbours, and drops off the more others are removed from them in social networks. So those who score high on happiness, have happier immediate neighbours in terms of twitter interaction, than those who are 2 or 3 links away, whose good cheer declines the further away from a very happy person they are.
Large sources of happiness on twitter, also seemed to have more interactions with extended networks. These hefty sources of happiness and ‘friendship’ were found to use words such as “you,” “thanks,” and “lol” significantly more frequently, while those who tended to interact less and who also tended to be less ‘happy’ were found to deploy negative words such as “damn,” “hate,” and “tired” more.
The authors of the study investigated alternative theories for this finding for cheerfulness to cluster around the very happy, and also to spread through the social network. Basically the glad tend to also know and interact with positive others. One possible explanation is a tendency for similar words to be exchanged between people. Using statistical methods, similarity of word usage was shown to not account for these findings.
But just in case you think the secret to happiness is just start interacting with hundreds of twitter users, another result from this analysis of 40 million tweets is that once you get near 150 others you interact with, it appears impossible to keep up regular meaningful social contact.
The authors point out the famous ‘Dunbar number’ of 150 seems confirmed by this research. Robin Dunbar is an eminent anthropologist who posited that 150 is roughly the maximum number of relationships it’s possible to pragmatically maintain.
This theoretically fits the number of people we would have encountered in our evolutionary past. Hundreds of thousands of years ago, throughout our lives, we evolved to inhabit small villages or communities of around 150 people. 150 appears a kind of fundamental limit of our social universe and there are even some theories that our brain cortex size has evolved to cope with this number and no more. Indeed one theory of unhappiness is that our modern world has become too socially complex and as a result we suffer. For example, we no longer live in communities which help us achieve the magic Dunbar number.
The authors of this study caution that it’s still not clear from their data whether happy people just tend to find each other and cluster for that reason, and the negative likewise, or whether happiness spreads like a virus. However previous research has supported the contagion theory.
For example, the famous Framingham Heart Study in the USA followed up 4739 people from 1983 to 2003 (before Twitter) similarly found happy and unhappy people tend to cluster, and happiness spreads up to three degrees of separation – to the friends of one’s friends’ friends.
James Fowler and Nicholas Christakis from Harvard Medical School discovered those surrounded by happier people and who are central in a social network, are more likely to become happy in the future. These clusters of happiness appeared to arise from the spread of contentment. A friend who lives within a mile and who becomes happy increases the probability that you will be happy by 25%. Interesting the same effects don’t occur between co-workers. The effect decays with time and with geographical separation.
We also know from other research that obesity, alcohol use and smoking appears to be influenced by social networks, by who you know and what they are up to. This week there is increased interest in the idea eating disorders might be spread through social networking.
In the future we might face faster more rampaging epidemics of psychological dysfunction, because of the instantaneous and pervasive nature social networking.
If we should increasingly regard emotions as kinds of infectious diseases, does this mean we should screen more carefully who we mingle with, and even quarantine some?