The influence of collaboration on research quality: Social network analysis of scientific collaboration in terrorism studies research groups

Abstract

Considering that counterterrorism measures are costly and cannot be afforded to fail, it is imperative that these policies are informed by evidencebased analysis. Yet the terrorism studies knowledge base continues to be dominated by conceptual pieces. To increase the number of evidence-based studies available to decision-makers, it is critical to test what kinds of research collaboration communities are the most likely to produce evidence-based research. The aim of the study is to fill this need. Drawing on models of knowledge production and social network analysis literature, normative and structural characteristics of a knowledge production system are delineated and operationalized as measurable indicators that are the most likely to correlate with the greatest amount of evidence-based studies per research community. A list of hypotheses is developed and tested in regards to terrorism research published between 1992 and 2013 by the means of regression and cluster network analyses of collaboration networks of authors and organizations contributing to terrorism research. The findings demonstrate that heterogeneous and organizationally diverse research groups are more likely to generate a greater amount of evidence-based research as compared to more homogeneous groups. Other predictors of the increase in the number of evidence-based studies produced by a research group are the group’s total degree centrality and internal link count, indicating that scientific collaboration among authors and organizations is essential to producing policy research based on data analysis.