Wednesday 23 March 2011

Over 100,000 Twitter users were psychologically assessed - without their knowledge or consent?


Social networks tend to disproportionally favor connections between individuals with either similar ordissimilar characteristics. This propensity, referred to as assortative mixing or homophily, is expressed asthe correlation between attribute values of nearest neighbour vertices in a graph. Recent results indicate thatbeyond demographic features such as age, sex and race, even psychological states such as “loneliness” canbe assortative in a social network. In spite of the increasing societal importance of online social networksit is unknown whether assortative mixing of psychological states takes place in situations where social tiesare mediated solely by online networking services in the absence of physical contact. Here, we show thatgeneral happiness or Subjective Well-Being (SWB) of Twitter users, as measured from a 6 month record oftheir individual tweets, is indeed assortative across the Twitter social network. To our knowledge this is thefirst result that shows assortative mixing in online networks at the level of SWB. Our results imply that onlinesocial networks may be equally subject to the social mechanisms that cause assortative mixing in real socialnetworks and that such assortative mixing takes place at the level of SWB. Given the increasing prevalenceof online social networks, their propensity to connect users with similar levels of SWB may be an importantinstrument in better understanding how both positive and negative sentiments spread through online social ties.Future research may focus on how event-specific mood states can propagate and influence user behavior in“real life”…….We collected a large set of Tweets submitted to Twitter in the period from November 28, 2008 to May 2009.The data set consisted of 129 million tweets submitted by several million Twitter users. Each Tweet contained aunique identifier, date-time of submission (GMT+0), submission type, and textual content, among other information…We complemented this cross-section sample of twitter activity by retrieving the complete history of over 4 millionusers, as well as the identity of all of their followers. The final Twitter Follower network contained 4,844,430users (including followers of our users for which we did not collect timeline information). Armed with the socialconnections and activity of these users we were able to measure the way in which the emotional content of eachusers varied in time and how it spread across links. [Happiness is assortative in online social networks,Johan Bollen, Bruno Gonçalves, Guangchen Ruan, & Huina Mao,March 2011]

The 4,844,430 users (whose tweets were scanned in 2008 for the aforementioned study) would work out at about eighty per cent of the estimated 6 million Twitter accounts in existence during that year. Although only those who posted at least one tweet daily over a six month period were retained in the study, which ended up formally assessing 102,009 users.

One of those harvested appears to be a co-founder of ISP/telco Sonic.net who happens to be ‘followed’ by Barack Obama.
The wife of a co-founder of SDMOMfia.com was caught up in the tweet trawl and she is also followed by the U.S. President.
Obama was found again following a PR person from the Detroit area whose name cropped up in connection with the study.
One male who has no public Twitter bio is yet another who is identified by this study. He links back to Obama through an account the President follows.
However, it is highly unlikely that Barack Obama’s official tweets were assessed for Subjective Well-Being aka general happiness as he didn’t tweet daily.
There are a number of other tweeters mentioned in the study who can be easily identified by the general public or by their own followers.

What on earth was Indiana University's School of Informatics and Computing thinking in allowing any persons used in this study to be identified either directly within the text or in presentations undertaken later by one of the authors (and in one instance assigned an emotional state)? Did no-one realise the power of Twitter and Google to disclose identities to the idly curious?

These tweets may be in the public domain, but surely there are limits to the uses to which others may put them.

No comments: