The findings of our study suggest that the doctoral mentorship relationship may play a significant role in student research performance in terms of both output and impact. The models explain 5% of the variance in productivity and 23% of the variance in impact, which indicates that while choosing the right advisor may positively influence one’s academic achievements, that decision alone does not tend to make or break one’s research career. The doctoral experience cannot be reduced to the advisor-advisee relationship. Additional factors, such as evolving in an intellectually stimulating environment, affect performance – more in fact than the quality and frequency of one-to-one interactions with advisors. These different degrees of mentorship relationships, such as coordination, cooperation, and collaboration, are not captured in our data. The results also suggest that providing co-authorship opportunities may be a good way for advisors to support their advisees, as it helps increase their research output and impact. Collaboration may be facilitated by students sharing research interests with their advisors, as suggested by the positive relationship between productivity and the similarity of the PhD dissertation and the advisor’s past work. The benefits of working closely with one’s advisor can conflict with the importance of establishing one’s independence as a researcher. Giving students the liberty to diversify their research interests may ultimately provide greater career advantages as they evolve in their careers.