Understanding scientific collaboration: Diversity of collaborators and sequences in publication bylines
Comparing distinct dimensions of diversity and author order in publications could improve the understanding of collaborative work and help to identify opportunities for and risks of evaluation of collaborative work in different scientific career stages. Thus, the aim of this paper is to better understand the contextual dynamics of evaluating collaborative work by investigating the relationship between the sequence of authors in the publication bylines and the diversity of their collaborators. The diversity of collaborators is quantified with three separate dimensions, namely topic, impact, and gender diversities. Using the ArnetMiner dataset containing ACM-indexed publications in computer science, we find that the following three patterns relate to higher-impact scientific publications: (1) greater (less) topic diversity of collaborators plus more (less) tendency to work as first or last authors; (2) less (greater) impact diversity of collaborators plus more (less) tendency to work as first authors; and (3) greater gender diversity of collaborators plus more tendency to work as last authors. We also detect different patterns of authors’ first-, middle-, last-authored publications before and after their Ph.D. graduation and find that there are some differences for topic and impact diversities but not many differences for gender diversity in terms of Ph.D. students and researchers in their later stages of careers. Some implications are described, and a brief description of the limitations and future research are provided.