The application of exponential random graph models to collaboration networks in biomedical and health sciences: a review

Abstract

Collaboration has become crucial in solving scientific problems in biomedical and health sciences. There is a growing interest in applying social network analysis to professional associations aiming to leverage expertise and resources for optimal synergy. As a set of computational and statistical methods for analyzing social networks, exponential random graph models (ERGMs) examine complex collaborative networks due to their uniqueness of allowing for non-independent variables in network modeling. This study took a review approach to collect and analyze ERGM applications in health sciences by following the protocol of a systematic review. We included a total of 30 studies. The bibliometric characteristics revealed significant authors, institutions, countries, funding agencies, and citation impact associated with the publications. In addition, we observed five types of ERGMs for network modeling (standard ERGM and its extensions—Bayesian ERGM, temporal ERGM, separable temporal ERGM, and multilevel ERGM). Most studies (80%) used the standard ERGM, which possesses only endogenous and exogenous variables examining either micro- (individual-based) or macro-level (organization-based) collaborations without exploring how the links between individuals and organizations contribute to the overall network structure. Our findings help researchers (a) understand the extant research landscape of ERGM applications in health sciences, (b) learn to control and predict connection occurrence in a collaborative network, and (c) better design ERGM-applied studies to examine complex relations and social system structure, which is native to professional collaborations.