DREaM event 2: Introduction to social network analysis
Dr Louise Cooke presented a workshop session introducing social network analysis at the the first DREaM workshop on 25th October 2011.
View this presentation on Slideshare
View this video on Vimeo
In opening her introduction to social network analysis, Cooke conceded that many people can be put off this particular method by the level of maths involved. Social network analysis derives from graph theory and mathematical sociology, so it can be complex mathematically, but there are now many software tools which take a great deal of the maths work out of the process.
She began by introducing the workshop participants to some of the key concepts and language of social network analysis. She defined dyadic ties, which are the links between a set of “actors” or “nodes” within a network. These ties may or may not have a direction. Social network analysis is focusses on the relationships between entities, rather than the properties of those entities, so the nature of the ties themselves are important. She acknowledged that the use of the word “social” often leads to confusion by those who assume that it involves analysing Facebook and Twitter, but that interest in the method has been piqued by the notion of the “six degrees of Kevin Bacon” which shows that everyone in the movie industry can be connected to Kevin Bacon within six links.
Cooke provided a context for social network analysis, which has been found to be particularly useful in knowledge management and for studying the communication patterns within the workplace. She explained that we have come to a realisation about the strength of networks and the access to ideas and opportunities that can be available through networks. However, not all of this is immediately intuitive. She highlighted the strength in weak ties, whereby the most successful job hunters have lots of weak links with people in different areas, rather than a lots of strong links within a smaller network.
Cooke discussed the properties of a network that can be examined, including whether the network is directed or undirected; the density of the network, which gives you an idea about how well the network is connected is a whole; and the centrality of the network, which focusses on the position of a particular actor within the network. She emphasised that social network analysis is the start of an analysis of a network, rather than the end of the process. Once the network has been mapped one needs to explore this and interpret the meaning behind the patterns shown in the network.
Cooke went on to discuss the practical uses of social network analysis in business, where companies often use it to help analyse their business after a merger to identify how information flows between employees. She highlighted the individual roles within a network can this can help to identify, including information bottlenecks, information brokers, boundary spanners and peripherals.
Cooke concluded the formal section of her presentation with a discussion about the practical aspects of data collection, including questionnaire surveys, interviews, observations, and analysis of existing datasets. She also discussed some of the software tools available to handle the maths aspects of analysis – including UCINET, NetDraw and NetMiner – and the need to consider ethical issues when presenting your findings using the graphs produced through these tools. She emphasised that data cannot usually be collected anonymously, but can be presented anonymously. Relationships are sensitive, so presenting data which maps relationships needs to be done considerately.
Finally, Cooke introduced the workshop exercise, which involved participants completing a questionnaire to help analyse to identify the properties of the network in the room. Participants were asked to indicate whether they were aware of the knowledge or expertise of others in the room prior to the event, and whether they had any direct acquaintance with individuals in the room prior to the workshop. She presented the preliminary results from her analysis of the first of these questions later in the afternoon.
The results showed that the network was quite well connected around certain key individuals prior to the workshop. There were twenty cliques or clusters of at least three people within the group, with the same names appearing in a number of these clusters, demonstrating a very centralised element to the network which could potentially be spread out through the project. The DREaM project co-investgators had a strong role within the network: Hazel Hall had the greatest degree of outward centrality, indicating she was the person who knew about the knowledge or expertise of the most people within the group, whilst Charles Oppenheim was the individual with the greatest degree of inward centrality, indicating most people in the group knew about his knowledge and expertise. Further findings are available on Slideshare and embedded below:
If you would like to comment on this presentation, please join the discussion in the DREaM online community.