Understanding the formation and dissolution of interdisciplinary teamwork networks: A comprehensive framework study of network structure, subject characteristics, and link attributes

Interdisciplinary research teams play a crucial role in addressing multifaceted problems that transcend the boundaries of disciplines. Despite the growing interest in these teams, previous studies have not simultaneously examined the factors affecting their formation and dissolution. In this paper, medical informatics is selected as the research field, and based on the team collaboration network in this interdisciplinary field, the comprehensive effects of its intrinsic structure, subject characteristics and link attributes on the formation and dissolution of collaborative relationships are empirically investigated using the Exponential Random Graph Model (ERGM) and the Separable Temporal Sequential Exponential Random Graph Model (STERGM). Results indicate that within network structural attributes, the presence or increase of closed triangular formations will significantly promote the occurrence or increase of the establishment of new collaborative ties, whereas open triangles impede formation and encourage dissolution. In subject characteristics, interdisciplinary of members positively influences collaboration formation, the enhancement of academic influence will increase the possibility of formation and decrease the possibility of dissolution, and academic productivity uniformly promotes both formation and dissolution. Additionally, scholars of varying academic ages exhibit distinct collaboration patterns. Within multidimensional link attributes, organizational proximity, disciplinary proximity, and collaboration history enhance formation and deter dissolution of collaborative ties, whereas topic similarity exerts an opposing effect. While previous studies have examined team formation or dissolution from one single dimension, this study proposes a comprehensive framework that integrates network structure, subject characteristics, and link attributes to elucidate the mechanisms of interdisciplinary team formation and dissolution. Also, by using a combination of ERGM and STERGM models, we can reveal how the aforementioned factors enhance team stability and reduce the risk of dissolution from both static and dynamic perspectives. Based on our findings, we propose targeted management strategies for different stages of team development, focusing on disciplinary balance, talent recruitment, and cross-disciplinary communication to ensure the team's sustainable and effective operation.