Anecdote: Social network analysis principles

« Anecdote circles | Main | Comments problem fixed »

3/03/06 |

Social network analysis principles

By Shawn Callahan. Follow me on Twitter. Filed in .

Social Network Analysis is more ‘social’ than ‘analysis.’ It’s true that data is fed into network analysis tools to create interesting charts but the value comes from people interpreting these charts and trying to understand what they mean. Making meaning is a social process and we’ve found in our SNA work that it is important to have some principles to guide SNA activities.

Here are 4 principles we use. Would love to hear of others you think are important.

use SNA for good not evil: the natural tendency when looking at SNA charts is to find your own name, or your group's, and compare it with other people and groups in the organisation. This can quickly lead to comparisons of things like one group having more hubs than another etc. Comparisons like this (from a measurement perspective) are unhelpful and should be avoided. It is better to ask yourself what connections would be healthy and see if they exist or look for structural issues (e.g.. no links between clusters) and then devise ways of helping people get into these structural holes. It's of foremost importance that the SNA never becomes a performance measure (implicitly or explicitly) because this will result in the technique ceasing to be a useful indicator of what's really happening—people will game the survey and fear the results.
don't jump to conclusions: there are many reasons why a link is missing or why a person might not be as big a connector as expected. Rather than jump to conclusions, use the observation to investigate further and seek to understand what is really happening.
engage from the outset those who will interpret the SNA results: people need to be prepared for the type of information they will see and what it can be used for, its limitations and understand their role in the sensemaking process. It’s important to get people who will interpret the charts on board from the beginning.
treat SNA data sensitively: without the background knowledge, such as understanding the role the original question plays in interpreting the results, SNA charts can be misinterpreted. Furthermore when people's names are visible levels of interest increase—and stories begin to be told. SNA data should be viewed as confidential and the charts and their interpretations should remain with those people with an understanding of the technique.

These 4 principles help guide actions and help get the most value from social network analysis. In a more complex world as things become more connected we will need tools like SNA to make sense of what’s happening. If not used well, however, the tool will become sullied in the organisation and become unavailable for future use. The way in which SNA is used at the outset will set in train the culture of how SNA is used in organisation.

Send this entry to: