Like Father(s), Like Son(s) : Does the Relation between Advisor and Student Productivity Persist on Group Level?

In light of the fact that scientific collaboration is increasingly wide-spread, we analyze whether the positive relation between supervisor research productivity on the one hand and young researcher productivity on the other does in fact persist in research groups where several PhD and postdoctoral students are supervised by a team of cooperating senior researchers. Our empirical analysis is based on a data set of 86 research training groups from different disciplinary fields funded by the German Research Foundation. We find support for our basic hypothesis that the positive relation between supervisor and student productivity also holds on group level. Concerning potential control variables that become relevant on group level, we analyze whether the productivity of students in research training groups is influenced by group composition with respect to different demographic variables (student age, gender and cultural background) and with respect to students’ study background. We find evidence for experience and gender effects; cultural and study background diversity, however, have no effect. Our results prove to be robust to a whole set of additional control variables such as group size and advisor-student ratio. Only when student satisfaction with the supervision of their advisors is included in the analysis, the relation between supervisor and student productivity is no longer significant – hinting at the more productive advisors being perceived as the better supervisors