The New England Journal of Medicine reported last week that obesity spreads through social networks. After the initial wave of headlines, I checked it out and was really impressed by the research, done by Nicholas A. Christakis and James H. Fowler.
One small (but to me significant) jewel of their research is the network movie they use to demonstrate their findings. See Judy Breck's post on SmartMobs for a link to the movie, which uses state-of-the-art network visualization algorithms to present complex data in an intuitive way. The authors used SoNIA to make the movie. (More on network movies here.)
Another gem accompanying this study is the editorial by Albert-Laszlo Barabasi. He points out the multiple levels of networks at play, which include not just social networks but also biochemical networks. Compare Barabasi's diagram to Peter Gloor's multi-level analysis of business innovation (at right) and you can see why there is actual scientific substance behind work such as Verna Allee's which "uses principles of living systems theory to help companies evolve management thinking."
Publicity like this makes it all too easy to go overboard with network science, however. Breck's otherwise solid post concludes, "A key truth the article demonstrates is that social networking is not some oddball result of the emergence of the Internet. Networking principles are deeply imbedded [sic] into the physical and psychological venues, as well as the virtual." I feel rather confident that the authors of the study would be puzzled by Breck's statement; a scan of their paper shows that they are confident enough in the foundations of network principles to use them in their work without any need to get defensive about their analysis.
What strikes me most about the study is the off-hand comment at the very end of the movie. We naturally expect obesity and non-obesity to cluster, and they do; however, as the authors note, we also see that obesity takes over the core of the social network and non-obesity moves to the periphery: Is this significant? My hunch is yes, and that there are implications for our health in this core-periphery structure.
This work is licensed under a Creative Commons Attribution-ShareAlike 2.5 License and is copyrighted (c) 2007 by Connective Associates except where otherwise noted.