Social capital remains an ambiguous term. The last decades it is used often, and often for different meanings. Recently, Chris Jones also mentioned it and raises great questions. Sometimes I also refer to social capital, but I learned to use the term with care, or at least explain what you mean by it when you mention it. You could avoid the term completely because of the ambiguity, but I prefer to keep using it.
Because it is being used for multiple meanings, but always related, I like to use it by combining at least two versions of the term. For example, The fact that it can refer to the social capital of a person (whatever exact definition you would give it) and for a group, makes it a multilevel concept. Somehow, the social capital of all persons in a group combined and the social capital of the group seem to be similar, but it’s really not. Making the step from one level to another must be done with great care.
This can be the basis for new challenges in research. Oh, Labianca and Chung (2004, 2006) did a great job with this challenge. In a way, because a group is made up of people, the individual social capital of these people is related to the social capital of the group. Under what conditions can the group perform? What is needed in terms of closure in the group, and bridging with other groups? What is the role of the individual people that are member of a group? What about people that belong to multiple groups?
In my current research, I focus on behavior of people in online groups. I look at people who form bridges between groups, distinguishing between the number of people that form the same bridge. We found some solid results there, which creates questions about what this means for the group or groups. So perhaps I need to make the step from the individual to the group, and probably social capital will be included in such research.
Below you will find my presentation for the Sunbelt 2013 conference in Hamburg last month. These are my first baby steps in the world of social network analysis research, and to me, (group) social capital is still a holy grail somehow…
This week I’m attending my first conference as a PhD candidate. It’s called Sunbelt 2013, and because it is in Hamburg and it rains non-stop, it better could be named Rainbelt. Nonetheless, this is the place to be this week for network researchers. And as I’m studying social networks, I’m glad to be here as well.
As a newbie, I can see it’s definitely a place where people meet old friends, make new friends and where it’s fun being part of it. Especially when on the Thursday, Friday and Saturday nights the drinks are included.
Yesterday and today I attended a workshop on dynamic networks and the software program ORA presented by Jürgen Pfeffer from the Carnegie Mellon university. A great tool, and when you are looking at temporal data, I think it’s essential. Probably a tool I can use in the near future.
Starting this afternoon, and continuing until Sunday morning, there are sessions where everyone can present their work. There are so many parallel sessions, that you will miss most of them because you have to choose carefully which ones to attend.
And as far as going into town and see what Hamburg has to offer, there is limited time for that as the Sunbelt program is very full, and I have to prepare for my presentation on Saturday morning as well. I will talk about people who bridge online groups by being an active member in multiple groups (which is slightly different from what you read behind the link). The data for the research comes from the online community we support at one of our clients from Favela Fabric, where I work.
At the picture below you see my view at the moment, it looks kind of silent and peaceful, but it’s getting busier every minute. Now I will continue to prepare my presentation while I still have some time…
In social network analysis, a network often looks quite simple, when you zoom in to a certain section. People are nodes, and they are connected to other nodes. Sometimes a connection means friendship, but it could also refer to advice giving, dislikes, knows, etc. etc. I’d like to see a connection as “works together with”, and nodes as people in a certain context, for example a large organization. In organizations nodes often belong to a cluster of nodes resulting in closed networks or clusters where ties are strong, and some nodes connect clusters, making them brokers. These ties are less strong. These networks are often visualized as a snapshot in time for sake of simplicity, and more often than not are overly simplified for readability purposes. Where people in organizational settings used to be member of one or two work-groups, nowadays with the rise in online collaborative spaces this membership is much more dynamic and volatile. Membership is much more voluntary than it is designed, groups emerge and dissolve faster and easier, and resources (knowledge, skills) come more from members themselves instead of the organization. Especially in the online world, albeit in organizational settings, this is and will be the case more often (well, in knowledge intensive organizations that is).
The above results in people being member of more groups than they were before. This can be as a core member in one or more teams, and it can be in the periphery in other teams. In social network terms this results in overlapping communities. There appear many bridges not made up of two different people (nodes), but a single node is forming a bridge by being a member of two or more communities at the same time. This is coined as a “structural fold” by Vedres and Stark (2010) as opposed to a “structural hole” coined by Burt (1992). To me, the “structural fold” is in abstract terms comparable to quantum mechanics. Where atoms in quantum-land can switch positions instantly (well, not exactly, but it can appear that way), people can too, when working with online collaborative tooling. It is common for many people to work at more than one project at the same time, dividing their time on different projects, not always knowing beforehand where to work on at what moment. That makes it possible to bring in knowledge and situations from one project to another almost instantly and by the same person. In network visualizing, there is a world to discover here. When a person connects two groups by being a core member for both, visualization could be relatively easy with Venn-diagrams. However, with more simultaneous multiple group memberships, and with more nodes in the network showing the same behavior, visualizing would be very challenging. I found the image below that illustrates what I’m referring to. The majority of the nodes are member of more than one group at the same time. With these numbers the visualization is good to interpret, but with growing numbers this will be a problem. Try to visualize overlapping communities with more than 10.000 people and hundreds of communities.
We see nodes being connected to other nodes, and being part of multiple groups. In this simplified example it is quite easy to interpret. In a global and large organization this would be quite problematic. Maybe when we add dimensions things would become easier. However, when introducing the quantum behavior as I just mentioned would introduce new difficulties when visualizing. Perhaps we have to let go of a person being a single node, a person can be many nodes at once. Person 1 can be at different ‘places’ simultaneously, and when a person is in which position is unknown, and perhaps irrelevant. The same is the case for person 2, 3, … n-2, n-1 and n. Showing and integrating their networks would be a great challenge. Maybe we can learn from current quantum visualizations. Nodes circling or jumping through network space via hidden dimensions. Although I wouldn’t be too happy when the controversial string theory would enter the social network space… Bottom line: a picture tells a thousand words, but that’s not always enough.
I’ve been reading literature about social network analysis (SNA) lately for my research. A lot is written about SNA. About analyzing, about measures, about SNA in organizations, and many more. However, many research does not address the value of the potential (or past) changes in the network, it especially addresses the value (social capital) of a snapshot of the network, the value of the existing social ties. Think about measures like density, distance, centrality, bridges, structural holes and weak ties, or, more qualitatively, trust, norms, power and autonomy.
In my view, it is not complete to study social networks as static. They were formed sometime before the analysis, there are reasons it became that way. It’s a bit like the universe, it changes continuously with changing nodes, relations and meaning of the relationships between nodes. Analysis of a network is always a snapshot in time. How the network will or can evolve is at least as interesting and important, because that will determine a snapshot at a later moment. The (social) reality that we live in now, is determined by the earlier realities, and the current reality will influence the possible future realities. Therefore we cannot deny the dynamic nature of a social network.
Now, can we determine the possible future directions of a network, for example in organizations? Can we identify what determined the current state of a social network? With the advance of online communities, and the vast amount of recorded data of relations and communication between people, perhaps we can. What network characteristics in the past influences the current network as it is? We can look at new entrants (nodes) which brings new opportunities, new knowledge, new relations over time. We can also look at nodes that disappeared (left the company) or changed position (got promoted). We can look at changing goals of individuals, departments or the company, and we can look at changing outside conditions (legislation, competitors, drastic events). There are many more things we can look at.
Not everything that we can identify in a social network snapshot is because of chance or fate. We probably can point to events in history that influenced the current state of the network. People made changes in their network themselves, or outside events triggered changes. Events can also be gradual, like the growing of a particular group within a network, that caused some change elsewhere in the network, which is an important asset in the current state.
So I’m thinking about looking for social capital measures in dynamic social networks, in the context of organizations, by comparing multiple snapshots in the past. It can hopefully be used to explain how networks work, and how they can evolve by making possible scenario’s, and what is needed to go for a preferred scenario. Do you think this would be interesting?