Scientific collaboration within Australasian cities

By Shaun Hendy 26/04/2010 1


Does scientific collaboration depend on city size?  And if it does, are smaller cities with fewer institutions and fewer scientists more collaborative?  Or do bigger cities with more specialisation and more opportunities for interaction support more collaboration?

Auckland 2009 To get at this question, I looked at scientific papers published in 2009 listed in the Thompson Reuters Web of Science database that had at least one author in a major Australasian city (Sydney, Melbourne, Brisbane, Perth, Auckland, Adelaide, Canberra).  From the list of co-authors for each city, I constructed the corresponding co-authorship network.

The 2009 Auckland co-authorship network is shown on the right.  In the middle sits the largest connected component of co-authors which contains 72% of the authors in the diagram.  Not all of these authors will be Aucklanders of course — many are in the network because they have collaborated with Aucklanders.  For example, I am in the network (somewhere in the middle)  because I co-authored an article with a colleague from the University of Auckland last year.

In a blog post last year, I constructed the co-author networks for the New Zealand CRIs using the same database.  What I found surprised me:  in 2008, more than 50% of CRI scientists (including me again) were connected through the largest connected co-authorship network (up from about 12% in 1994).  I also looked at the 2008 University co-authors and found that 70% of them could be connected by a single network.  So Auckland looks pretty connected.

To put those Aucklanders in context however, let’s compare them with other major cities in Australasia.  Below I’ve plotted the number of co-authors associated with a selection of the major cities versus the proportion of those co-authors in the largest connected component. Australasia 2009
Auckland is actually at the low end of the scale, along with Perth and Canberra.  At the high end, the largest components in the Melbourne and Sydney co-author networks occupy close to 90%.  Larger cities do seem to exhibit more connectedness amongst researchers.  If you accept connectedness as a proxy for collaboration, the big cities in Australasia were more collaborative in 2009.

Interestingly, when you put New Zealand itself on the plot, you find that it is more connected than Auckland.  This will not surprise many south of the Bombay Hills!


One Response to “Scientific collaboration within Australasian cities”