Social network metrics are for eggheads

Thanks to Frank van Ham, co-creator of ManyEyes and researcher at IBM's Visual Communication Lab, for giving me the title of my next talk: "Social network metrics are for eggheads." I don't know where or when this talk will occur, but that will be the title.

One case study I will share as part of the talk will be the the following wonderful project by Valdis Krebs. He was working with a group advocating for healthy affordable housing. Time and again they had cornered one dilapidated building or another into cleaning up its lead paint, only to hear the owners announce, “We just sold the building, sorry.” After hearing this excuse a few times, they sensed something was up. With a little research at the library, they found all the real estate holding companies involved in the series of sales were associated with one “godfather” who ran the whole empire through a network of children, siblings, spouses, and in-laws. The piece of evidence that ultimately convinced the judge to throw the book at the lead-paint mafia was a network map like so:
Real estate companies are green at the top, godfather/mother are at the very bottom, and intermediaries who own/run the companies are in the middle.

Those of us participating in the debate over how best to measure social networks in ways that predict important outcomes (such as revenue, profit, or shareholder value, to name a few) can forget that there are also simpler and less controversial applications of SNA. Valdis' Family Ties picture above is one great example.

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.

Guide to effective network survey design by Terry Murphy

Terry Murphy has written an outstanding (and sobering) assessment of online network survey tools. He has very kindly allowed me to post his report. The abstract is below:

This research report takes questionnaire usability best practice and online usability in a questionnaire context as the basis for developing usability guidelines for social network analysts who choose to employ online modes of data acquisition.

Highlighting the distinct character of network analysis questionnaires, and the opportunities, and potential cost efficiencies, of acquiring network analysis data through online networks, the relationship between usability and communication in the context of questionnaires, particularly the importance of visual design, is considered.

The report provides eleven usability guidelines for online network analysis questionnaires ranging from the general to the network-analysis specific.

The guidelines are not new or revolutionary, but rather provide a network analysis-specific overlay to general online questionnaire guidelines.

Three freely-available, low-cost online survey applications are assessed against the guidelines.

You can get the full report here.

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.

Harnessing the power of networks

In celebration of the third birthday of Connectedness (officially Friday but I can't wait that long) and in tribute to the power of networks, I want to honor the dozens and dozens of unhappy Sears customers who have turned my blog into the #1 Sears customer relations website. That's right, if you Google "Sears customer relations" there is only one website ranked higher than mine, and it's the Sears Investor Relations site--a mistake, clearly--so I like to think that my site is the true #1 Sears customer relations site on the Internet.

How did this happen? It's all about "harnessing the power of networks," baby! One day I posted a story about a bum refrigerator (with a happy ending no less) and three years later that post has risen through the Google ranks, gathering comments with ever-increasing network power. Now I am host to 31 tales of Sears customer woe (and counting). By the way, that does not include the emails I get regularly on this topic.

Within this experience is an important lesson in the reality of "harnessing the power of networks." No matter how well I learn this lesson, though, it fails to soften the sting of knowing that my broken-refrigerator story is the most popular page on my blog by a factor of two. This does not feel like it is burnishing my SNA-guru credentials.

Still, it could be worse. Thanks, everybody, for not posting anything illegal on my site (unlike Digg).

Connectedness year #4 promises to be our best, most network-harnessed yet!

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.

New network movie algorithm

Shortly after hearing my "Complaining about Network Visualization" talk at the University of Maryland, host Ben Shneiderman tipped me off to this award-winning paper presented at EuroVis 2007, where he had just given the keynote address. Ben's tip directly addresses my complaint that most SNA tools consider visualization as an isolated problem to be done once. The problem is that we like to look at networks change over time and watch series of network maps. Generating series of network maps that communicate meaningful information is no easy task. The movie below by Yaniv Frishman and Ayellet Tal summarizes their significant step forward in automating this task. (15MG QuickTime)
Here's an overview of my numerous related posts on network visualization.

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.

LINKS: Intra Organizational Networks in Business

Big news from the Gatton College of Business and Economics at the University of Kentucky, which very recently founded LINKS, the International Center for Social Network Research. The homepage says

The mission of the center is to bring together scholars from different academic disciplines who share a common interest in social network research and application. Our overarching goal is to conduct and publish cutting-edge research in the rapidly expanding field of social network analysis; and to serve as a bridge between the science of social networks and real world organizational problems. The core insight of the network perspective is that the pattern of relations among the elements of a system—be they people, organizations, neurons, or computer servers—has important consequences for the system's performance.

Some very impressive people have moved to Kentucky to join Dan Brass, including Steve Borgatti, Joe LaBianca, and Ajay Mehra. Keep your eye on LINKS.

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.

Qualitative Data, Quantitative Analysis

Pacey Foster (soon to be professor in the School of Management at UMASS Boston) points me to this essay by H Russell Bernard, "Qualitative Data, Quantitative Analysis." It presents a balanced overview that takes the qualitative vs quantitative discussion into separate dimensions of data and analysis. So instead of provoking an argument that pure number-crunching cannot completely describe multi-cultural organizational effectiveness (as I did yesterday), Bernard invites consideration of the following two by two matrix:
What does this mean for analyzers of social networks? We work in all four cells. In my experience we tend to move through them roughly in sequence:The only real beef I have with Bernard's discussion is where he says, "strictly speaking, there is no such thing as quantitative analysis of qualitative data." Since he wrote the essay in 1993, before the "digital revolution," I will let him off the hook for suggesting such an impermeable boundary between these two.

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.

Technology and humanity

For a summer-time philosophy stretch, I recommend Robert Pirsig's classic novel, Zen and the Art of Motorcycle Maintenance. The last time I read it must have been at least ten years ago, when I was deep into abstract theories of computation. Now I am reading it with new eyes. The book is largely about the boundary between abstract "explaining" and physical "experiencing," a boundary tied closely to "quality."

The book should be required reading for anyone who applies for the Department of Defense grant, Assessing Performance of Complex Organizations. This grant seeks "a method of assessing organizational performance that minimizes human input or intervention." Furthermore, "The resulting tool will account for [cultural] diversity as it is actually represented within the organization being modeled and assessed."

I am both thrilled and horrified by the thought of assessing, with as little human interaction as possible, multi-cultural organizational performance. I want to participate in this research agenda, partly to build the model that it seeks, but also to question the premise on which it is based.

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.

Maryland HCIL and SocialAction

My complements to Adam Perer and Ben Shneiderman for hosting a great workshop, "Helping Users Make Sense of Social Networks." They are warm hosts and sharp researchers. I enjoyed being one of their designated "users."

All the slides from the workshop are posted here. These include my previously mentioned complaints about SNA software. They also include the latest from Adam's PhD research project, SocialAction, the most intuitive network visualization tool I know. Since I last saw SocialAction, Adam has added annotating capability not unlike Word "revisions" mode--allowing several people to collaborate over a series of potential edits. My mind is racing with possible applications. In case you're curious, SocialAction is not publicly available (yet). Hopefully soon.

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.