I had been toying with the idea of updating my collaboration network for some time when I happened upon a new R package ('scholarnetwork') that does most of the hard work. The last time I looked at my collaboration network, I focused on how my collaborations evolved from grad school to my post-doctoral experience, with a particular emphasis on geographic location. Since then, I have developed new collaborations while also strengthening existing ones, leading me to wonder if and how the features of my collaboration network have changed.
To do so, I extracted my collaboration network from my Google Scholar profile using 'scholarnetwork' and then visualized it with 'ggnet' . I then weighted the links between myself and each co-author by dividing one by the number of co-authors per publication, and then summing these weights per co-author. This emphasizes co-authors with whom I publish more frequently and generally with fewer co-authors. I also scaled node and label size (where each node is a co-author) by the node degree, which is the number of co-authors each co-author is connected to in my collaboration network. This highlights co-authors that I have collaborated with on publications with large numbers of co-authors. I then grouped co-authors by affiliation and in two cases by either a synthesis project ('Stability') that has been a major focus of my post-doctoral experience or by collaborations/initiatives that I am involved with (e.g., TRY). Lastly, I organized the layout of the network using the Fruchterman-Reingold algorithm to make it look slightly less like a ball of cat hair. My current collaboration network, I think, shows that my post-doctoral experience is now at the core of my research profile, yet still bears the imprint of my work during grad school. Most of my collaborations as a post-doc are concentrated in the bottom half of the figure, either in bright red or in dark blue. These connections are also linked with collaborators in the 'Other' category, which reflects that most of my work as a post-doc was focused on synthesis projects that often involved international working groups. Another noticeable feature of the bottom half of this figure is that strength of my links with most of my co-authors is fairly weak, which again emphasizes the collaborative nature of my postdoctoral research experience; a majority of my publications as a postdoc have involved more than 20 co-authors. In contrast, the upper half of my collaboration network - dominated by my research as a grad student - involved fewer co-authors. Because of this, my direct links to my PhD supervisors are stronger than in other parts of my collaboration network. I have the general feeling that this is more typical for PhD students based in the US than those based in Europe or China, where PhD students are typically involved in larger research projects. This part of the network also reveals how later I became connected, through my PhD supervisors, to a productive research network focused on secondary tropical forests (see links connected to F. Bongers and R. Chazdon). Another feature of my collaboration network that I find interesting are the nodes located at the periphery (in the lower left, far right, and upper left of the figure). These nodes represent publications that I have become involved with because of data that I have shared. While exciting to be part of these large collaborations, I feel that their value to my development as a scientist has diminished with time. Initially, it was a valuable experience to see how these manuscripts evolved and how the lead authors managed the review process. Now, I am thankfully at a stage where the benefit of being involved in such collaborations is more marginal and I would benefit more by spending my energy, time, and ideas on my core research. Over the past several months, I have spent more time than usual thinking about my research program: what lies at the core of what I am doing now and where I want to it go in the future. To a much greater extent than writing a job application (and less pressure), I have been pleasantly surprised by how useful a tool building my research collaboration network has been to clarify my thoughts on the future direction of my research. |