The influence of academic advisors on academic network of Physics doctoral students: empirical evidence based on scientometrics analysis

Abstract

Scholarly socialization is a crucial and fundamental component of doctoral training in preparing future scholars. One of the important goals of doctoral student socialization is to build up the academic network, that is, to be loosely defined, to develop connections with other scholars in knowledge production and exchange. Academic advisor, as the mentor and guide for doctoral students through their journey entering the academic world, plays an important role in helping students establishing their network. Using bibliometric data on the publications of recent doctoral degree earners in Physics, this study examines the influence of their advisors on the formation of their academic network. Specifically, we randomly sampled 1% of the doctoral degree earners (thereafter called “new Ph.D”) in Physics who graduated from a university/college in China between 2001 and 2018 and whose doctoral dissertation was available in the CNKI database by July 31, 2018 (1022 people in all). From the dissertation, we were able to grasped the name(s) and institution(s) of the new Ph.D’s academic advisor(s) and the publications he/she published during doctoral training. Then with scientometrics analysis of the publications, coauthors, and citations, we established the co-authorship network and citation network for each new Ph.D, and estimated the structure of networks (number of nodes, density and centralization) and the position of the new Ph.D in the networks (in- and out-degree, centrality and constraint). The primary interest of this study is to examine whether and by how much these two features are influenced by the academic ability of the advisor(s). The study further examines whether the supervision model (i.e., independent guidance, joint guidance, or team guidance) may influence the features of the new Ph.D’s networks. Preliminary findings suggest that Doctoral supervisor, by introducing Ph.D. candidates to academia with his or her own academic contacts and influence, could make the Ph.D. candidates acquainted with academic elites, thus accelerating the academic socialization process of the Ph.D. candidates. Comparing to single supervision, the team supervision model is more helpful in establishing the co-authorship network. However, the doctoral supervisor has an extremely limited influence on the position of the Ph.D. candidates in the academic network, and what’s more, the Ph.D. candidates have not become a bridge for scholars to seek cooperation and transfer knowledge. There is no difference between the models in establishing the citation network. The above findings provide implications for academic supervision in the training of doctoral students’ research ability to enhance the contribution of their research products.