Proximity matters for collaboration

The last few years have seen a plethora of digital tools created to allow for virtual collaboration, but a recent MIT paper suggests that physical proximity is still key when supporting collaboration among academic researchers.

The study examined over 40,000 published papers and 2,350 patents from MIT researchers over a 10 year period, and mapped out a network of collaborators across the university, before then examining the locations of each collaboration, particularly in relation to the departmental and institutional membership of each researcher.

“Intuitively, there is a connection between space and collaboration,” the researchers say. “That is, you have better chance of meeting someone, connecting, and working together if you are close by spatially.

The Allen Curve

Central to the finding was the so called Allen Curve that was devised by MIT professor Thomas Allen.  It suggests that collaboration diminishes as a function of distance.  Indeed, even simple conversations are significantly less likely to occur when people are over 10 meters apart.

Despite Allen’s work first appearing in 1977, and a large number of digital tools emerging since then to try and mitigate its impact, the MIT paper suggests its findings still hold true today.

MIT is a good place to conduct such research as most buildings on campus house many different academic groups.  This scattering of areas of inquiry means that the impact distance between workspaces has on collaboration can be clearly measured.

Interestingly, the results differed depending upon whether the output was papers or patents.  For papers, researchers located close to one another were three times more likely to collaborate than those located just 400 meters apart, with this becoming stronger the further researchers were apart from each other.

For patents however, the curve was less steep, albeit still prominent as researchers in the same building were twice as likely to collaborate as those 400 meters apart.

“There is a persistent relationship between physical proximity and intensity of collaboration,” the authors note.

The team believe that the study could be extended to take in both more campuses but also more time to study how collaborative activity changes, especially as the location of researchers changes.

Either way however, they believe that their work adds a nice architectural dimension to the field of scientometrics, and as such provides a nice input into decision making around workspaces and campus design.

“You can never predetermine what research will be novel and powerful and exciting,” the authors conclude. “But you can create the conditions for collaborative innovation to happen.”

Related

Facebooktwitterredditpinterestlinkedinmail