Methods and Models for Co-Authorship Networks

Scientific collaboration, recognized as a crucial driver of research progress and innovation, has increased significantly across all academic disciplines. This trend is further reinforced by government policies at both national and international levels, which actively promote collaborative research initiatives. In this context, co-authorship serves as a tangible manifestation of collaborative behaviors among scholars. While research topics and methodological approaches often differ between disciplines there are communities that share common ground. This is exemplified in Italy by the coexistence of Economics and Statistics within the same macro research group, as well as very often within the same department in many Italian universities. This proximity suggests shared similarities in departmental and university environments, as well as alignment with national strategies and policies regarding scientific production and research quality. However, key questions arise regarding the potential convergence of scientific production mechanisms between these two communities. Specifically, does this shared environment influence co-authorship behavior, shaping co-authorship structures, publication style, and productivity over time? To address these questions, this contribution aims to conduct a comparative analysis of co-authorship networks in Economics and Statistics, starting from their network topology and modeling the dynamics through the Relational Hyperevent Model (RHEM), a family of statistical models that explain the propensity of a group to co-participate in a future event (such as a paper) given the participation in past events.