About SNA (not NSA)

In my blog in April I wrote how Joitske, Josien and I were familiarizing ourselves with SNA, social network analysis. We continued our exploration and on the 8th of November we presented three cases for discussion to a selected audience of network professionals.

Social network analysis visualises connections between people or organisations. With SNA software large amounts of data can be gathered and processed. Hence my reference to the NSA that, as we know by now, is certainly also using social network analysis to monitor interaction on a large scale. But, of course, for proper use of SNA data should be collected with consent. We used surveys to get our information (with the software ONASurveys).

Before we worked on the cases, we were fairly unlettered in SNA and frankly quite sceptical if it was worth the time investment. All three of us have a background in using participatory facilitation methods ‘under the mango tree’ in developing countries. We believe that data should be owned, and made sense of, by the users themselves and not be presented as facts by external experts.

From this premise it will probably not surprise you that we prefer to use SNA mapping as input for discussion and reflection by the network members themselves. Also after applying SNA on the cases I sustain this opinion, but I am convinced now that SNA adds a very useful perspective and provides important new information.

One case we studied was de Linkedin group LOSmakers. It was launched in 2012 to discuss the use of social media for online learning and accompanied by an excellent publication. When discussing a number of network graphs with the facilitators, Joke and Lyset, it surprised them that some people are more active than they had expected. Further they observed that that some of the initiators (the authors of the LOSmakers publication) are becoming less involved, whereas others who joined during the last half year are gradually taking a more prominent role.

Graph 1: Time of membership of the group. Some members who joined only in the last half year (yellow) are already well known contributors and move to the centre.


The facilitators were happy to see renewal in the network. A number of more active people who are motivated to invest in the network could be identified and the SNA exercise helped them to find strategies for increasing involvement of new people.

They also agreed that the next two graphs clearly show that the network is building social capital.

Graph 2: Red lines new connections of people who met because of the network. People connected with black lines already knew each other.





The LinkedIn group LOSmakers has a clear objective to stimulate exchange, so a good question to ask was “from whom participants had learned”.

Graph 3: Lines point at people that at least 2 others say to have learned from.


Joitske had the opportunity to use SNA as extra activity for two assignments with large open networks, and likewise the results were discussed with the coordinators of facilitators of the networks concerned. (I helped with some additional analysis in one case.) Although also in these cases the maps gave some insights, Josien who is involved intensively in the networks raised doubt if the SNA exercise would be too much a meta-analysis. She argued that the two networks consist of many pragmatic do-ers who make effective use of these networks, but who are not concerned with the overall picture (the networks is just a given and it works for them).

When we presented the cases to the meeting of network professionals, Josien’s observed that SNA has limited use for an open network that is growing and changing continuously. An interesting and somewhat heated discussion followed. One point debated was that SNA is pointless in one case of open network (revolving around European agricultural policy) because the issue is too broad and there was no clarity on the purpose of the networks. What do the links between people in the network mean? Are they just connected because they know each other from working in the same field, or are there specific activities or topics that they contact each other about? The main question to be asked was: should you first be clear on what binds people before you produce maps of the network?

I feel that different interesting aspects were mixed in the discussion:

  • First of all, it needs to be clear for whom the analysis is made. The users are likely to be the people involved in coordination of the network. I see no problem if not all network members will be interested in reflection on the functioning of their network.
  • Small demarcated or large open networks will look quite differently when mapping them with SNA. So you will use SNA to look at different aspects. In smaller networks coordination roles might be more important, whereas in more fluid networks the response on one ore more particular issues could be of interest for agenda setting.
  • It matters what type of connections your network graph shows. With the same people you can produce a different map of their connections if you ask with whom they relate professionally or whom they like to invite for a drink. Before using SNA you need to know about the key activities in the network. These activities can change over time, so the survey design is not fixed and might well include different questions next year. Setting up a good survey is an important skill to learn before doing SNA.

Before the presentation, we asked our audience if they considered SNA to be complicated, too technical or distracting from more important issues. The opinions were somewhat reserved at first, but gradually more positive. As long as you are careful not to draw too definite conclusions from the graphs alone it was seen as a useful addition to understanding networks.

We concluded that SNA:

  • Helps visualise what is often already known or suspected. It helps to make explicit if members understand their network in the same way.
  • Is useful to discover persons who are active and involved. Some names surfaced that were a surprise to the network facilitators.
  • Contributes to discussing the further development of the network and take strategic decisions.
  • Has most impact if the analysis is done jointly with representatives of the network.
  • Does not take much time if (online) surveys can be used.

There are evidently many more ways of applying SNA, but our first exploration was successful and inspired to more. My particular interest is to discover specific roles in the network like brokers or coordinators, which did not feature clearly in these cases. Further I would like to use SNA to look at information flows in network. Therefore my search will continue…