On blogging and collaboration

We’ve submitted a paper to Frontiers of Ecology and the Environment that deals with the art of collaboration in large-scale ecological research.  It’s in review at the moment, so I’m not going to talk too much about it, except in setting up my discussion here.

Jacquelyn Gill has a new post up that talks about the roles of writing, blogging, getting papers out and submitting grant proposals.  One comment she includes is that she has received advice indicating that when push comes to shove, blog posts don’t count toward tenure.  It’s an interesting comment, on one that I suspect comes from someone who doesn’t blog.  While I agree that blogging isn’t going to matter much as far as a direct benefit, I think it plays a strong role in fostering collaboration.

Figure 1.  My publication network. I am the dot in the center, collaborators are linked by lines based on co-authorship.  The plot is coded using R.  Code used is included here.
Figure 1. My publication network. I am the dot in the center, collaborators are linked by lines based on co-authorship. The plot is coded using R. Code used is included at the end of the post.

One point we make in the Frontiers paper is that collaboration is not an immediate benefit of research.  It is a mid-term benefit of research projects.  You need to accomplish something, and it needs to be evaluated, before people are willing to reach out and ask, or willing to accept collaborative proposals.  Aside from my direct Ph.D collaborations (set up through my advisor), and a great opportunity from Jana Vamosi, my first real collaborative projects didn’t happen until after my first paper was submitted and I presented my work at the IPC/IOPC in Bonn, Germany.  It was a great conference all around. I met a number of researchers there who were doing great work, and I began working with Odile Peyron.  But, as I mentioned, this collaboration only happened because I was able to submit work early in my Ph.D, and because my advisor (Rolf Mathewes) was able to send me to a great conference and help support me in my career.

This gets me back to blogging.  A number of researchers have used the social web to incredible advantage.  I know whenever Jacquelyn Gill tweets my posts I get a huge increase in traffic here, in part because the people who follow her blog trust that she is tweeting or blogging relevant material that will interest them.  In the same way, I hope that my blog is getting my name out and building a form of trust with its followers.  It certainly appears to be:  My direct research products have collectively been accessed over one hundred times through their DOIs, and the number of hits my posts get is steadily increasing.  So, aside from the social aspect of blogging, I am directly familiarizing people with my work beyond the standard conference forums.

If blogging isn’t counted toward tenure, at the very least it helps you make connections to your discipline and your work year round.  When people meet you at a conference they have something to discuss, and that familiarity helps the development of collaboration in the long term.  One of the important points about collaboration is that, without familiarity, the types of stresses inherent in a research project (time, money, interpersonal issues) can be magnified.  It’s no surprise that Wright and Bartlein (1993) credit time on the University of Wisconsin Terrace for helping to ensure the success of the COHMAP project.  That kind of familiarity helps build trust among collaborators.

By providing a public outlet for your ideas, you are helping foster familiarity with your point of view, and participating in broader outreach to a trans-disciplinary peer group.  In the mid-term this should help increase the likelihood of collaboration.  I have reached out to blog writers with skill sets complimentary to my own in the interest of collaboration, but I have no other concrete data on how pervasive this is.  Maybe your blog has led to collaboration?  Let me know in the comments.

Finally, here’s the code for the figure above.  Replace the bib file with your own and let me know what you find!  If you have any questions, let me know.

#  Author network mapping for journal publications.
#  By: Simon Goring (final code: 10/2/2013)


#  Read in the bibtex file
#  You can export this file from your google scholar network.
#  Ultimately, this code can be used with any bibtex file though.
citations <- read.bib('SJGcitations.bib')

#  Process the author list, first create an n x n matrix with all the authors
#  then go paper by paper and increment when authors co-occur.
authors <- lapply(citations, function(x) x$author)

unique.authors <- unique((unlist(authors))[names(unlist(authors)) == 'family'])

coauth.table <- matrix(nrow = length(unique.authors), ncol=length(unique.authors),
                       dimnames = list(unique.authors, unique.authors), 0)

for(i in 1:length(citations)){
  paper.auth <- unlist(authors[[i]])[names(unlist(authors[[i]])) == 'family']
  coauth.table[paper.auth,paper.auth] <- coauth.table[paper.auth,paper.auth] + 1

#  Build the network diagram:
author.net <- network(coauth.table)
network.vertex.names(author.net) <- rownames(coauth.table)

col.set <- brewer.pal(6, 'Dark2')


#  ?plotnetwork will explain these commands
aa <- plot(author.net,
           label = rownames(coauth.table),
           usearrows = FALSE,
           jitter = TRUE,
           displayisolates = FALSE,
           vertex.cex = log((colSums(coauth.table))),
           edge.col = rgb(0.1, 0.1, 0.1, 0.1),
           label.cex = 0.7,
           label.pad = 0.7)

Published by


Assistant scientist in the Department of Geography at the University of Wisconsin, Madison. Studying paleoecology and the challenges of large data synthesis.

9 thoughts on “On blogging and collaboration”

  1. Very cool post! I love the idea of coauthorship as a simple network plot. It is really intuitive and effective. I also like your comments about blogging….expanding our networks is never a bad thing, even when new relationships don’t lead to (immediate) collaboration on pubs.

    1. Thanks Andrew. I’ve been following your blog for a while as well, glad you’ve started posting more lately, I think when I first found it there had been a bit of a lull, but the code examples you provide are great! I have to work through the googleVis one sometime soon!

  2. hi,
    I am extremely grateful for this blog – I was looking for an example like this for a long time. Finally I ws able to create co-authorship network entirely in R (you can see the result on our webpage: http://www.i-deel.org/publications.html , at the bottom).

    This is a very nice and simple code and is very well annotated. However, I had two little problems running it.

    First, it required bib files that have key fields, and then I had to modify this line: unique.authors <- unique((unlist(authors))[names(unlist(authors)) == 'family'])
    to this: unique.authors <- unique((unlist(authors))[grepl('family', names(unlist(authors)))])

    Second, sizing vertices like this: vertex.cex = log((colSums(coauth.table))), does not represent the numbers of publications coauthored with the person in the centre of the graph (thats what I wanted to show), so I have modified it to: vertex.cex = log(diag(coauth.table))

    Hope thats useful,
    Thanks again for sharing the code,

    1. Thanks! What a fantastic diagram too! I love the little clusters in there, but it’s very nice and clean. I’m glad that the code (mostly) worked for you, and that the changes were relatively simple. Obviously one of the real strengths of R: it can be scaled, and it’s reproducible and adaptable 🙂

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s