It's very easy to become enamoured with the visualisations and potential of network analysis, and see it as an end unto itself. This is one reason why I think that network analysis is a diagnostic methodology. It can aid understanding, but there are obvious limitations. For example any visualisation is a representation, or report of, data collected at particular time in a particular place. We all know human systems are dynamic, so it's reasonable to assume data will degrade and the network will change.

It is also a common mistake to think the visualisation, or the data matrix, represent analysis: they do not! I think Drew Conway , who is a political scientist, makes the point very well when he says: ...

read more ...

Now here's an interesting challenge. The editorial board of the Journal of Social Structure has announced the first annual symposium on network visualisation. The theme is "Visualisation in the Service of Understanding". The board is challenging network researchers and practitioners to produce a single figure and caption that tells a compelling scientific story in a way that tables or text alone could not. I'm always trying to do this and call the resultant visualizations SOILS . A SOIL is a senior officer interest light, which goes on when they are attracted to a diagram, graph, map or picture. The SOILS Syndrome is a propensity to privilege these visualisations over other forms of knowledge, often at the expense of proper understanding, but I digress. ...

read more ...

I'm often asked what tools I use to collect, record and visualise data. I almost always collect data using a semi-structured interview technique. I find the results are richer and allow me to explore areas that a web-enabled survey won't allow. Similarly, I almost always record the answers directly into Microsoft Excel, although this isn't always practical. Quite often I build a database in Microsoft Access and manipulate data as required. My analysis and visualisation tools of choice are UCINET , NetDraw , and NetMiner 3 . Each has its strengths and weaknesses, which I might deal with in another post. Increasingly, however I find myself using NodeXL ; at least to do the hack work.

NodeXL is a free Microsoft Excel 2007 add-in. I blogged about NodeXL some months ago when it was called ".NetMap", however I understand for legal reasons the name was changed to NodeXL. NodeXL is now at build 88 and just gets better and better. It now features zoom and scale, more metrics, better ribbon layout, more filtering options, additional layouts, and better control over the graph pane contents. ...

read more ...

I've been reading a good deal about program and portfolio management of late. I even did the United Kingdom Office of Government Commerce's Portfolio, Programme and Project Offices course a week or two ago. This course builds on the PRINCE2 methodology for projects, but I digress. One of the problems I find with portfolio and program management is the lack of simple high level tools to measure the health of the organisation. RAAAKERS™ Profiling, which I'm developing as part of my PhD, may provide a tool.

RAAAKERS™ stands for Responsibility, Authority, Accountability, Awareness, Knowledge, Experience, Resources and Systems and is as a way of representing the main attributes associated with management of a large or complex enterprise. A bottom up approach is applied by collecting data from end users and then aggregating it into a visualisation. I've used it a few times and my colleague Doctor Mark Burnett has also used it with success - see this peer-reviewed article for a description of its use in an organisation. Anyway today I thought I would look at data collected in my PhD and see if the resultant visualisations might be of use. Consider the graph below.

read more ...

I had dinner and a few wines with a good friend on Friday night. He is interested in network analysis and wanted to read some materiel that is grounded and relatively easy to understand. I'm also having dinner with a PhD student from ANU on Thursday night. I have no doubt the same topic will come up, although he will want something more substantial. Coincidentally Professor Barry Wellman and Professor Alexandra Marin , have just published "Social Network Analysis: An introduction ", which is a draft chapter for an upcoming book on social network analysis. It's very readable and a succinct introduction, and one that I would highly recommend to anyone.

Of course it's very hard to go past the resources available at INSNA , and particularly Connections , which is a peer-reviewed journal and the definitive source for reliable grounded work. A recent issue of Methodological Innovations presents some good quality articles on social network analysis. The articles on both sites can be downloaded for free.

Another very interesting site with a number of excellent papers is Learning for Sustainability . They have a whole section on building and mapping networks. ...

read more ...