Skip to main content

Why Social Network Analysis?

Submitted by: Zoe Madden-Wood on Mar 28th, 2013 at 5PM PDT

In the new era of leadership, we have learned that collaboration and networked leadership models trumps individual models of heroic leadership. We’re wiser and capable of greater things together than apart. But how do we measure if we’re succeeding at collaboration? If we’ve created a healthy network? Or if we have gaps that we need to close in communication?

This is where social network analysis comes in. Social network analysis is a type of analysis that measure networks of people and helps evaluators determine how people are connecting and around what issues and projects. With social network analysis, you can take a snapshot of the network and figure out both the network strengths and weaknesses, and use that to grow a better and more robust network for a greater and more dramatic impact.

What does a network look like?

Most of the time we use network maps to make meaning of the work. Below is a map of respondents and non-respondents in blue and green respectively. As you can see, there are three separate networks that are very strong, with one of them being much larger than the others. Between these networks are various “bridgers”, i.e. people who are bridging the two groups.


Basic Network Terminology

  • Node
    A node is a person, organization, or idea. (These are the dots in red.)
  • Link
    A line that connects two nodes. (These are the lines that show connections.)
  • Cluster
    A cluster is a group of nodes that are more connected to each other than to other nodes outside the cluster. (Those are highlighted in yellow in the clusters.)
  • Hub
    A hub is a node at the center of a cluster. (This is the green star.)
  • Bridger
    A bridger is a node that connects different clusters. (This is the purple star, spanning the clusters.)

Adapted from  Bruce Hoppe and Claire Reinelt, Social Network Analysis and the Evaluation of Leadership Networks, Leadership Quarterly 21 (2010) 600-619.


Below is a the growth of a disconnected network into a healthy network. It starts with scattered clusters and then becomes a strong web of interconnectivity.



Network maps can be a great tool to move people into action, which is why it’s always good to get a sense of people’s interests and what they’d like to partner around.

What’s involved in a network survey?  

The difficulties with a network analysis is that for best data, it requires a 75% response rate. This is because you’re not trying to sample the data and extrapolate the whole based on a smaller subset. No, you’re trying to actually map the whole network. And while if you’re missing a few connections here or there, it may not change things much, if you’re not getting a high response rate, it’s hard to definitively say anything about the network.


What this means for your campaign?

It means you have to be very aggressive about getting people to fill out the survey. If you have a conference that many of these people are attending, showing up with paper copies is standard practice for a network survey. Calling people and emailing them individually is also a good idea. And this means at a minimum that you have to be sure that your email addresses and contact info is up to date for everyone on your list.

Survey - Types of Questions

The survey goals in a network survey are tied to the goals and needs of the network. Most of the survey can be your typical survey. What’s your name? What are you interested in? Where it gets interesting is that we’re also asking about relationships with people. We display a list of people and we’ll ask between 2-4 questions about that relationship. In general you want your relational questions to be specific actions within a certain timeframe. So, you’ll have a list of Sam, Phillip and Kevin, and after that will be questions like, “Have you called, emailed or met with this person within the last 6 months?” This will generate the “links” in the network maps.

Because this is a cutting-edge technology, special software is needed both to take the survey and to analyze the data. Survey Monkey and like-minded applications don’t often have the survey data in the right format. They don’t produce the relationship outputs need for SNA. Instead we use software such as June Holley’s software, NetworkGenie, and onasurveys.

Once we have the data, special software is necessary to generate the maps. UCINET, NodeXL, and InFlow are some of the major networking softwares out there. There’s also Gephi as an open-source option.

In general, SNA is a great tool to both measure your network and to activate your network.

Want to learn more about Social Network Analysis? We have a variety of webinars and articles we’ve done on this work, including Network Analysis Methods for Assessment and Measurement You can even have fun and explore your own network (a personal network is an “ego network) on Facebook by going to Wolfram Alpha.


Inline image 4