Social Science Based Research at NCRC: an interview with Felix Kabo

August 14, 2013

1.Please tell us why NCRC is interesting to your research which is in the social sciences:

NCRC is a great opportunity to understand the effects of spatial networks on research collaborations. Over a year ago, we started working with pilot data from the BSRB and the Cancer Center. At that time, with fewer people at NCRC, we did not have a robust enough data set. However, over the last year, a number of groups have moved to NCRC, including ones which were not previously co-located before. In fact, we could make the argument that some of the groups would not exist but for NCRC. Now with the availability of more research groups and data that we can begin to collect, we plan to study the collaboration process at NCRC. Often when you look at collaboration it is difficult to disentangle the effects of space from the effects of the organizational hierarchy and other effects such as social networks, and relationships.

At NCRC there are groups such as the IHPI which consists of smaller groups of researchers but the institute had not existed before as a unit. For us it was interesting to think of their move to NCRC as “disruptive,” but in a good way. It can shake the existing patterns of relationships and once things settle, we can actually start to see the effects of moving. If we look at where collaborations and relationship were at time zero which is right before or right around when people move and then look at it again in the future - a couple years down the road, 5 years down the road, 10 years down the road, etc., we can start to see the effects of moving to NCRC. And because this is such a unique opportunity, we are very excited about NCRC.

One of the stated goals of NCRC is to encourage collaborations across boundaries, which is actually very hard to do, although U-M is better than most of its peers when it comes to creating the environment and frameworks for this. But it is still a daunting task. NCRC can function as an incubator or as a lab for these types of cross-disciplinary collaborations. Usually, in conventional set-ups, people are separated by organizational boundaries, buildings, etc. But at NCRC we can have situations where, for example, in building 20W, on a single floor there is a lab where on the one end is cancer or oncology research but the other end is basic chemistry research, with dentistry, chemical engineering and other fields in between. These are the types of co-locations that we are unlikely to find on any other part of campus. NCRC is creating opportunities for these kinds of interactions, making it more likely that people will form relationships and collaborations. And then out of these collaborations we are likely to see some qualitative or quantitative change in innovations.

2.Given the opportunity to study co-located research groups and diverse disciplines together at NCRC, how exactly are you studying them and what results do you hope to find?

At this time, we are actively seeking funding to study NCRC in greater depth. However our pilot study has implications for the NCRC study. The pilot was interesting - we looked at administrative data for the Medical School going back to 2000. From that we pulled subsets of people who moved to the BSRB in 2006. Additionally, we included people who were in the Cancer Center between 2000 and 2011. These two populations of faculty researchers were more similar, yet quite different from the rest of the Medical School population, making it interesting to study the impact of space on their research, including moves to new locations that happened for both these groups.

This means that we can start to see what happened to collaborations before and after they moved, including developing new methods of collaboration. Typically, people study research in terms of number of patents developed, publications, etc., which are good metrics but the problem is that they take a while to mature. For example, if you see a publication that is coming out this year, it has probably been in gestation for three years. As a result, we may miss three years’ worth of work done prior to the publication. We looked at IRB applications and grant proposals -- whether successful or not -- and created a composite index of collaboration. Then using this index, we were able to study the effects of space. We also created a new measure of proximity called zone overlap, where we defined the zones as the spaces that individuals effectively use, occupy or visit in the course of a regular working day, and the paths that connect them, hence path overlap.

What we found was to the extent that people’s paths overlap, it is not only predictive of whether they enter into a collaboration relationship, but it also predicts whether or not within a 3 year window their grant proposals get funded. The results were significant. Path overlap was significant in both the buildings we studied, so it suggests that this new measure is robust to changes in how the building is configured. This has direct implications for NCRC.

Our pilot study provides us with a great proof of concept for the NCRC project, especially since we have developed two new measures relating to collaboration and the effects of space.