Several studies suggest that Science-Industry knowledge exchange occurs via many links and complex interactions among organizations (innovative firms, academic and research institutions). Most of empirical applications have taken a simplified view of relationships among actors by focusing on a single type of tie at once. The aim of this contribution is to take into account the multiple ties that are at the base of knowledge and innovation diffusion on a local environment. Two kinds of interactions will be considered: co-authorship and co-patenting observed on individual level among scholars and inventors, affiliated to public and private organizations located in the area of Trieste (North-East part of Italy). Our main assumption is that knowledge transfer from university to industry can flows via different types of collaboration. To this end, a multivariate exponential random graph model (MERGM) will be proposed to jointly examine the multiple relationships structure among actors and their roles in the multivariate knowledge network. Multivariate modelling for network data is a rather recent approach, with relatively few applications. An additional aim of this work is therefore also to discuss the potential of this approach in organizational studies.
