Social Networks: Beyond the Numbers
Social networking research is becoming a hot topic as people discover the potential that mapping has for guiding policy and practice, however like many other “hot” research methods, there is a need to go beyond the numbers to make sense of what they really mean lest we create beautiful maps and have no place to go with them.
The rise of social media applications and the ability for anyone to use simple tools to create, extend and shape their social graph with a mouse click or tap of the app has helped stoke interest in network research. Social networking methods are those that tend to favour quantitative development of maps and numerical representations of what a social network looks like. We are most often terribly ignorant of the role and position we play within a network so to see ourselves and peers positioned literally on a map can be revealing in more ways than one.
Between 2005 and 2006 my colleague Tim Huerta and I did a study that looked at the formation of a community of practice (CoP) in tobacco control through the lens of the social network and published a paper on this based on that work. Part of the study involved giving a group of people who were meeting face-to-face a survey and ask them about who they knew that influenced their work in the web-assisted tobacco interventions (the CoP’s focus), how well they knew the people they identified, and what kind of things they did together. This data was entered over the first night and analyzed for the next day’s meeting with the network map revealed (see summary here, full article here) .
While the map itself was generated through quantitative analysis, revealing patterns akin to the image above, it was the meaning that people gave to those connections that allowed this group to begin envisioning how they could leverage their untapped potential in the network to advance their interests as a community. This sense-making process is too often neglected in social network research and risks turning something meaningful (like a relationship) into a statistic that can more easily be dismissed — or misunderstood.
This week another social network study was published looking at tobacco control and the use of Twitter as a medium to support that. This study didn’t map the network per se, but rather looked at the type of people in the network and what the content of the messages that were shared within it. This represents another type of social networking study where the researchers aim to peer into the activities of a group of people who are interconnected and describe from afar what they see and who they think they see. This has some utility for those wanting to delve into areas where there is little known (such as Twitter and smoking cessation), but can also mislead people if used improperly. Networks are dynamic, with influence shifting and participants activity modulating greatly within its lifespan and because of this, cross-sectional data poses the risk of capturing a slice of activity that might not reflect the whole.
Consider the analogy of slicing a watermelon sideways and doing so at the end, rather than in the middle: if you love watermelon, you want the middle slice because it’s bigger and richer than the end. The same might be true for social networking activity.
The tendency to want to produce network maps using numbers alone to explain them is highly problematic. Even Facebook has decided it will give greater priority to what people do in the network, rather than how big the network is as evidenced by the push to add Skype-powered video to its service this week. Facebook knows that their value is determined less by how many friends you have, but more about how you truly connect: photosharing, comments, game-playing.
True, this can all be quantified, but they are going a little further beyond the numbers. The Facebook example provides an interesting example of potential social network studies that could look at the type and content of the photos shared, how people have reacted to them, and what kind of social movement has been formed by the content created for and shared through that network. If you want to leverage a network for social change and good, this is the kind of stuff you need to focus on, not just the total numbers of people involved.
Powerful social network research is as much about having a good statistician as it is an anthropologist and together, they need to have the story that comes from it, woven together by the users and a good storytelling host.