The doctoral advisor, typically the principal investigator (PI), is often characterized as a singular or primary mentor who guides students using a cognitive apprenticeship model. Alternatively, the cascading mentorship model describes the members of laboratories or research groups receiving mentorship from more senior laboratory members and providing it to more junior members (i.e., PIs mentor postdocs, postdocs mentor senior graduate students, senior students mentor junior students, etc.). Here we show that PIs laboratory and mentoring activities do not significantly predict students skill development trajectories, but the engagement of postdocs and senior graduate students in laboratory interactions do. We found that the cascading mentorship model accounts best for doctoral student skill development in a longitudinal study of 336 PhD students in the United States. Specifically, when postdocs and senior doctoral students actively participate in laboratory discussions, junior PhD students are over 4 times as likely to have positive skill development trajectories. Thus, postdocs disproportionately enhance the doctoral training enterprise, despite typically having no formal mentorship role. These findings also illustrate both the importance and the feasibility of identifying evidence-based practices in graduate education.



Document Type




File Format

.txt, .csv

Publication Date



NSF, Division of Graduate Education


Utah State University

Award Number

NSF, Division of Graduate Education 1431234; NSF, Division of Graduate Education 1431290; NSF, Division of Graduate Education 1760894

Award Title

Collaborative Research: Progressions of Skill Development in Biology Doctorates; Collaborative Research: Progressions of Skill Development in Biology Doctorates; Trajectories into Early Career Research


In total, we recruited 336 participants from 53 institutions across the United States. Data for the present study were obtained through web-based surveys and the collection of single-authored writing samples via email. Both survey data and writing samples were collected annually during the first 4 y of the doctoral program. After removing cases with missing data on all key variables and accounting for attrition (both from the study and the doctoral program), the present study relies on a longitudinal sample of n = 297 students

Referenced by

Data set supports article:

Feldon, D. F., Litson, K., Jeong, S., Blaney, J. M., Kang, J., Miller, C., Griffin, K., & Roksa, J. (2019). Postdocs’ lab engagement predicts trajectories of PhD students’ skill development. Proceedings of the National Academy of Sciences, 116(42), 20910–20916. https://doi.org/10.1073/pnas.1912488116



Code Lists

see README file




Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Additional Files

README.txt (13 kB)
MD5: 160291585ef39cc513d3b2e5cad0a62d

pnas2019_postdoc_data.csv (130 kB)
MD5: 9f8d94f758bb36a7adba4cecffd5cd31