14542: Navigating the Frontier of Network Models: Advances and Limitations of Contemporary Approaches

Network research has become a topic of interest for most, if not all, sub-disciplines of management research – be it organizational behavior, strategy or innovation. Indeed, recent reviews unambiguously state that, for some fields, the analysis of inter- and intraorganizational networks has become a fundamental pillar of their discipline (e.g., Burt et al., 2013; Chen et al., 2022; Jacobsen et al., 2022), a continuation of a decades-long trend (Jarillo, 1988; McPherson et al., 1992). While much of our knowledge of organizational networks stems from the theoretical strides made by management scholars, methodological advances made by statisticians has also played an essential role in pushing the field forward (Hunter et al., 2008; Parker et al., 2023; Snijders et al., 2010;), perhaps more so than with other fields. A potential reason for this is that observations in networks are inherently interdependent and thus pose an analytical challenge for many ‘traditional’ statistical approaches, which treat observations as independent. Yet even though methodological opportunities are vast, a majority of research on organizational networks still employs linear regression approaches, thus risking a less-than-ideal theory-method fit (Carpenter et al., 2012). At the same time, many of the current developments in network models focus on improved estimation routines and algorithm efficiency, leaving those management researchers familiar with network statistical methods longing for additional modelling options to answer pressing questions, for instance about brokerage. Overall, there is an opportunity for management scholars to employ recent network statistical models and for the development of models to address new theoretical questions.