Research We Love


Spatial arrangement of social and economic networks among villages in Nang Rong District, Thailand

I love this early embedding of SN nodes into GIS. This Thailand study shows that tractor hiring networks occur over different land uses (rice vs. highland crops). it also finds edges’ distances from various third places like schools and temples to show how those institutions curate relationships. This is a fantastic, comprehensive example of how GIS can be used to enhance social network research.

Clustering of 770,000 genomes reveals post-colonial population structure of North America

How cool is this? Researchers took advantage of a bunch of geolocated DNA samples to tell us how lineage is spread out. It isn’t very often that we see the U.S. partitioned or demarcated by such linear strips (except for some highways), economically, socially, etc….this map takes the notion of tight, geometrically-sound clusters and turns it on its head.

The pitfalls of A/B testing in social networks.

The OKCupid blog is one of my favorites. Here’s an entry that maps some of their data.

Any geographer would have had an easy answer to the A/B testing problem. So eloquently put by, …. And voila, Lovesheds. The most interesting part of this, to me, is the division between North and South Chicago. Perhaps racial divides also divided the city into multiple dating zones?

Hinterland Boundaries of New York City and Boston in Southern New England

It asks a simple question: where does NY stop and Boston begin? It creates contour lines 10%/90%, 50%/50%, 90%/10% around New England using phone calls, check cashing proclivities, travel, newspaper subscriptions, etc, etc. It’s fun mapping in the 1950s, and its author was one of the VPs of a grocery store (Stop and Shop). I weasel this paper into almost all of my classes.