I've been playing with a new (to me) analysis tool called Gephi. It's really impressive, and was able to quickly show me that students are mostly looking at the profile pages of students in their year, and less often at the profiles of students in other years. Below are two different renderings of the same social network data. Apologies for the lack of a colour key, but each colour corresponds to one cohort of students. The keys are different, and it's actually the same cohorts in the same location on each graph, with the newest cohort at the bottom, and the oldest cohorts mixed in together at the top (that mixing is partly an artifact of my data collection - I only captured the last year or two of those cohorts' data, and in those years the students from different cohorts are more mixed in together, in their various hospitals.
Next steps: look at the detail within a particular cohort, and see what effect clinical school allocations, stage of the program, and PBL tutorial group have on the data.

