Existing theory suggests that social networks form a major factor in individual and team performance, including in academic collaborative research. However, there is currently a lack of a theoretical framework to explain the social network related factors that influence publication processes and decisions. We address this gap by adopting a theory‐building perspective, analysing a large data set on the impact of social networks. Using a case of one leading journal (Human Resource Management), we collected data on 10 years of publication, exploring 327 papers, written by 667 authors, which represent 839 ties. Examining individual demographic and institutional characteristics, we developed and tested a set of hypotheses, providing a framework for shaping and guiding future academic research collaborations. Our findings suggest that as albeit the intention, diverse collaboration is still at a low level. This contribution adds to the literature on understanding the factors influencing the type of collaborations that lead to publications in leading journals.