|
 |
|
Application of Social Network Analysis
The 7th Annual International Conference on Digital Government Research, dg.o 2006, will be held in San Diego, California from May 21 – 24, 2006. David Lazer, Director of the Program on Networked Governance at Harvard University, will chair a tutorial on "Application of Social Network Analysis in Digital Government Research," says Lazer. “Our objective is not to make you a social network expert—which is impossible in 3 hours—but to familiarize people with the key ideas and methods from social network analysis.” The idea of social network analysis dates back to the 1930’s and the pioneering research of Jacob Levy Moreno, who coined the term “sociometry” for his measurements of how closely people were socially related (see Wasserman and Faust 1994). One of the field’s most famous ideas is familiar to the layperson from the popular notion of “six degrees of separation.” It was inspired by the “small worlds” work of social psychologist Stanley Milgram. Although Milgram is infamous for his obedience experiments involving putative electric shocks, he also did equally significant and (far more benign) work on societal connections. Some connections are obvious: friends, family, work or academic relationships. But Milgram created an experiment in which people were asked to send a letter to a complete stranger, routing it through the acquaintance they thought most likely to have a connection to the stranger. Although the actual number of completed chains in his first experiment was extremely low, the average number of links in those chains was six. But the idea is far more complex than how many actors can claim a working relationship to Kevin Bacon (see http://www.cs.virginia.edu/oracle/), which is why researchers like Lazer find social networks so fascinating to study. For example, Milgram found that people’s perception of a letter’s importance affected their urgency to pass it on. Consider: It would be all too easy to conclude that a student had no social circle if you asked her to forward an email about an alumni picnic. But if you asked her to forward an email about a free U2 concert taking place on campus that night, you might find a stadium filled within a few hours of the original note. In a famous paper published in Nature in 1998, Duncan Watts and Steve Strogatz quantified the “small worlds” idea, allowing it to be applied to nearly everything from computer networks to Dutch Tulip Mania. Their work even kicked off a mini-genre in publishing, including Watts’s own popular book, “Six Degrees.” “I think it hit the zeitgeist at just the right time, just as the Internet was emerging, and that’s such a powerful metaphor for human relations,” says Lazer, “ I think the idea of networks generally has exploded over the last decade.” Another concept receiving much attention is “power laws,” the idea that networks are not distributed randomly, but cluster around powerful nodes, whether among people, computers or organic systems (see Barabasi and Albert 1999). A power law, Clay Shirky points out on his blog, may be contrasted to the better-known Bell Curve distribution. Due to the Bell Curve idea, we’re often used to thinking that the largest cluster should be in the middle of a distribution, but power laws say that a distribution will be strongly tilted, as well as affected by the choices of others. The idea is easily illustrated at fan conventions. Whether it’s country music or Star Trek, the largest clusters of fans will be found nearest the biggest stars. And those clusters will grow, as other fans think, “Look at the size of that line, it’s got to be for somebody big.” http://www.shirky.com/writings/powerlaw_weblog.html All of these, and many more issues in social network analysis, are pervasive, complex, and crucial, says Lazer. One important example, he says, is how the idea is approached in law enforcement. Police and legal professionals constantly need to consider clusters of networks, and how they may—or may not—interact. If all the members of a drug ring have the same dental office in common, does that make the dentists criminal masterminds? Or just doctors who are so good, all their patients keep recommending them to their friends? Another point to consider is “social capital,” the idea that relationships affect productivity. Both the dentists and the criminals increase their skills and output by associating with their peers. But if learning from peers in a network is the upside, getting stuck in one’s own professional silo is the downside. Social network analysis helps people see where the bottlenecks are, and how the very structure of organizations, and inter-organizational relationships, may help or impede achieving desired goals. “Government is often made up of entities that have to work together, but don’t always have authority over each other,” says Lazer, “In the field of public administration, there’s been an increasing realization that teaching students the world works in a hierarchical fashion leaves out a very important reality, which is that you have to work with other organizations that are equal to your own in terms of legal authority.” Lazer offers his workshop as a way for Digital Government researchers, whether academics or practioners, to learn more of these and other social network analysis theoretical frameworks, in order to inspire new ways of analyzing their own data, “The social network analysis paradigm is a lens through which people can pick out things in the landscape they otherwise wouldn’t be able to see.” Albert-Laszlo Barabasi, Reka Albert, Emergence of scaling in random networks, Stanley Wasserman and Katherine Faust. Social Network Analysis. Cambridge University Press, New York, 1994. D. J. Watts and S. H. Strogatz. Collective dynamics of 'small-world' networks. Nature 393, 440-442 (1998). | ||||||
|
This site is maintained by the Digital Government Research Center at the University of Southern California's Information Sciences Institute. |
|
• CONTACT • POLICIES | ||
| | |||||
