Institute for Software Research
School of Computer Science, Carnegie Mellon University


Collaboration in an Academic Setting:
Does the Network Structure Matter?

Victoria A. Hill

July 2008

Center for the Computational Analysis of
Social and Organizational Systems (CASOS) Technical Report


Keywords: Collaboration, social network analysis, dynamic network analysis, network structure, network metrics

What forms of collaboration result in the most benefit to individuals who are in the business of creating new knowledge? I approach this question by examining patterns of collaboration among university faculty members with the objective of determining what types of collaborative relationships are most likely to result in innovative ideas and knowledge creation. By drawing on the toolkits of Social and Dynamic Network Analysis, I measure different structural positions of the network of actors based on this collaborative behavior.

The dataset used in this study contains publication and collaboration data from 1995 to 2006 for each of 61 tenure or research track faculty members in the computer science department of a major U.S. university. Publication data was used as a proxy for knowledge creation. Co-authorship of publications and inter-departmental collaborations on projects, grants and students were used in calculating several network metrics including the E-I Index. These metrics along with relevant control variables are subsequently used in a multivariate regression model to estimate their significance on total publication rates of faculty members. Results indicate that innovation and new knowledge creation are facilitated by new inter-departmental partnerships for a specific cohort of faculty members.

19 pages

Return to: SCS Technical Report Collection
School of Computer Science homepage

This page maintained by