Peng Wang
Dr. Peng Wang is an Associate Professor of Innovation Studies at the Department of Managment and Marketing, School of Business, Law and Entrepreneurship, Swinburne University of Technology. His research focuses on the development and application of statistical models for social networks, particularly Exponential Random Graph Models (ERGM) for social selection processes and Auto-logistic Actor Attribute Models (ALAAM) for social influence processes. As the designer and programmer of the PNet suite of software packages, Dr. Wang has made significant contributions to the statistical modelling of complex networks, including one-mode, two-mode, multiplex, and multilevel networks, as well as the co-evolution of network structure and individual outcomes. Dr. Wang has publications in the research fields of public health, innovation management, social-ecological systems, anthropology, political networks and network intervention evaluations.
Sessions
In this hands-on workshop, participants will learn the fundamentals of estimating Exponential Random Graph Models (ERGMs) and Auto-Logistic Actor Attribute Models (ALAAM) with MPNet – a software developed to investigate the structural features of networks and how such structure may affect individual outcomes. ERGM for multilevel networks enable researchers to investigate the influence of structure at one level of analysis on structure at a different level, while taking into account the complex interdependencies that exist within and between levels. For instance, interpersonal networks between managers at the micro-level might interact with alliance networks of the organizations they are nested in. Throughout the workshop, participants will work through short exercises to get familiar with the graphical user interface and output of the MPNet software. We will discuss various case-study examples that will provide the participants with a good understanding of the possibilities that multilevel ERGMs offer for social scientists.
How do organizational members come to recognize someone outside their own area of expertise as a leader? Leadership identity research conceptualizes leadership as socially constructed through processes of claiming and granting, yet granting has largely been understood as arising through direct interaction between potential leaders and followers. We argue that this dyadic view becomes incomplete when leadership crosses expertise boundaries, where actors may lack the shared knowledge and evaluative criteria needed to assess one another directly. In such settings, advice networks may provide relational cues through which expertise becomes visible and leadership can be granted.
We examine this argument using whole-network data from 228 healthcare professionals involved in a multidisciplinary breast-cancer clinical pathway. We model leadership recognition as a directed network and examine its association with formal managerial role and direct and indirect advice relations using exponential random graph models. Our primary analysis focuses on leadership recognition across professional boundaries, with within-profession leadership providing a theoretically meaningful reference case.
Formal managers are more likely to be recognized as leaders across professional boundaries, but formal position does not account for the broader relational pattern of leadership recognition. Cross-professional leadership is strongly aligned with direct advice relations and, importantly, is associated with indirect advice pathways connecting actors through third parties. These pathway effects are substantially more evident across than within professional boundaries.
We contribute to leadership identity and organization theory by extending the granting of leadership from a predominantly dyadic process to a network-embedded one. Our findings suggest that, across expertise boundaries, leadership recognition is supported by advice structures through which otherwise difficult-to-evaluate expertise becomes visible.