http://www.sciencedaily.com/releases/2009/02/090213161031.htm

How can we measure and study social networks?  In real life, with ever-changing networks, this is nearly impossible. However, the creators of Second Life, a virtual social community, designed a way to do so. In Second Life, over 15 million users create online personas and interact with others.  Northwestern University professor Noshir Contractor is studying the networks on Second Life and also Teen Grid, the virtual world exclusive to teenagers.  With regard to the inherent freedom users have online, many worried that teenagers who participate in these online communities would develop friendships with strangers and “potential troublemakers.”

However, Contractor’s findings were just the opposite.  He discovered that teenagers had a larger number of online friendships “with people in their immediate geographic area – likely with people they already knew.”  Teenage users on Teen Grid were also likely to be friends with the friends of their friends, people who were probably a part of the network already.

Contractor’s findings support the triadic closure principle.  Of the three supporting reasons for triadic closure, homophily seems the most applicable here.  In online networks, people tend to become friends  through common characteristics, such as location or similar interests.  Since most of one’s Teen Grid friends were those who were geographically close, this suggests the presence of few actual bridges.  If many of the friends in a network already knew each other, then they would most likely have a dense social network with few actual bridges.  There would most likely exist other ways for a node A, for example, to access a node B’s group if the edge between the two was disconnected because they would have other friends in common.

Another reason for Teen Grid users to become friends with people they already know is social influence.  As seen in the analysis of the LiveJournal community, as the number of one’s friends who joined increased, there was a greater likelihood that that person would join the network too.  Although the marginal effect diminishes after a substantial number of friends have become members, once a second friend joins the network there is a larger influence to join as well.  One’s online friends will then begin with the people who introduced him/her to the network, that person’s friends.

Contractor is also studying what happens when Teen Grid users turn 20 and must exit the network.  They can now join Second Life, but they are suddenly forced to “leave their entire network of friends behind.”  Social influence suggests that if at least two friends join Second Life, the user will join as well.  Since people become members once they are 20, this would imply that a substantial amount of one’s Teen Grid and then Second Life friends are around the same age, especially if the friends are people they already know and who live in the same area. 

When Teen Grid users begin Second Life, is there the same threat of strangers that was a fear in Teen Grid?  Even if users have some apparent qualities in common, the other principles of triadic closure suggest that if the other friends don’t know the person, the user will not become friends with the stranger.  This should serve as some comfort to the parents who worry about their children connecting with strangers across the world online. Despite having similar interests, a potential “friend” of someone with no other friends in common makes users wary and decreases the likelihood that they will become friends. 

 It will be interesting to see what happens when Teen Grid users abruptly leave one network and join another, and Contractor will continue to study these online networks and compare them to those in real life.