This May Shane and I are teaching a new course on Social Network Analysis on edX. This was quite a challenge to design as I was reluctant to give into the linear transmission of small idea chunks, and instead produced materials that integrate skills and organize them by competencies thematically. I am quite proud of some parts of the course, but also am so well aware of all the things that are not, and that could go wrong. It is the first iteration though so I feel remind that to myself.
The course is organized in three weeks.
Week one focuses on explaining graph metrics – this week is most useful to those who want to know how to calculate centrality or density in the network or plot it. Knowing how to do these things does not fully help design and implement a network study, and the assignment of this week that tests learner understanding and application is only 15%. Instead of recording lectures, I created so-called extended notes. Week one includes many terms in the Extended Notes that commonly appear in network analysis studies. I thought of week one as the basics you have to know to do anything at all, the common terminology, the common definitions of things like nodes, ties, types of networks, density, degree, centralisation, clustering.
Week two is more about conceptual tools, ie how to use more network thinking, which questions to ask and which considerations to make when designing or conducting a network study. Also, week two is about presenting theoretical constructs about human behavior that are often tied to anlaysing network. This week talks about network questions, testing network hypothesis, ethics, as well as a handful of heavy-duty concepts such as social capital, homophily, preferential attachment, assortativity, strong and weak ties and brokerage. The annotated script gives examples of how some of these constructs may be operationalised through network metrics, but calculating the metrics is not even the part of the assignment. This week is critical as it is about thinking and concepts. The assignment tests one’s understanding of how to construct a network and the relationship between network metrics and network interpretations.
Week three is also about conceptual thinking but this time placing all that has already been introduced into digital data domains. How does one’s thinking about the network interpretation and operationalisation of network ties, edges and metrics change once we use digitally collected data (not self-reported one)? What still makes sense? What changes? What becomes limited?
I also designed the course so that if the learner engages into the experience of a week, there is a reflection to follow, and an input summary with key weekly take aways at the end. I am sure I had done enough to confuse the learners, hopefully just enough to balance their intent and not to turn them away. We will see soon.