The “Cascading Mentorship” Model: Postdocs Enhance the Training Experience of PhD Students

By Roberta Attanasio, STEMM Leadership Editor

Everyone benefits from mentors, although not all mentoring relationships are equally successful. There are many different mentoring styles, mentoring models, and mentoring relationships. Careers, now subjected to paradigm-shifting forces, are continually evolving—thus, mentoring styles must evolve with them. Mentoring relationships are successful when they are tailored to specific individuals and their interests.

In the sciences, active mentoring is essential to create a supportive research environment. However, the intense competition for funding, career advancement and jobs can limit the time that senior scientific leaders can dedicate to mentoring, thus contributing to the challenges experienced by younger scientists who aspire to become leaders themselves.

Photo by Christian Lue on Unsplash

For PhD students, mentors are often their doctoral advisors—the so-called principal investigators (or PIs for short). Typically, one PhD student has only one PI, and therefore one mentor. Often, the PhD student/PI mentoring relationship follows the “cognitive apprenticeship” model. In this model, a more experienced person assists a less experienced one, providing support and example, so the less experienced person gains new knowledge and skills. However, this model has been used throughout the years without any evidence that it is actually useful for the development of research skills.

Interestingly, results from a recent study (Postdocs’ lab engagement predicts trajectories of PhD students’ skill development) published October 15, 2019, in the Proceedings of the National Academy of Sciences, show that the PIs’ lab and mentoring activities don’t significantly predict the skill development of PhD students—rather, it’s the engagement of postdocs and senior graduate students in laboratory interactions that supports the advancement of PhD students. In other words, the “cascading mentorship” model is at the basis of a more vigorous skill development.

According to the “cascading mentorship” model, members of laboratories or research groups receive mentoring from more senior laboratory members, while at the same time providing it to more junior members—PIs mentor postdocs, postdocs mentor senior graduate students, senior students mentor junior students, and so on.

For the study, researchers used a longitudinal design. Over a period of 4 years, they measured the research skills of a cohort of 336 PhD students in the biological sciences (more specifically, cellular and molecular biology, and developmental biology) from 53 universities across the United States. The researchers found that, when postdocs and senior doctoral students actively participated in laboratory discussions, junior PhD students were over 4 times as likely to show year-over-year growth in all research skills that were measured—this is despite the fact that postdocs and senior doctoral students typically do not have a formal mentoring role in research groups.

The researchers note that “as the practice of science has shifted toward larger team enterprises and an increasing pace and volume of workload, the nature of the PI’s role has shifted to one that often entails less direct contact with students.” Thus, postdocs and others within the laboratory may provide the direct contact necessary for the skill development of junior PhD students. The researchers suggest that training postdocs in effective mentoring practices may further enhance the benefits to graduate students.

In addition, the study results indicate that the total value provided by postdocs within the laboratory may be substantially more than currently recognized. Notably, while contributing to the development of PhD students’ research skills through informal mentoring, postdocs may enhance their own skill development. Indeed, the researchers cite exploratory studies suggesting that, by mentoring students, postdocs not only further develop their own research skills, but also other important skills such as teaching and scientific communication.

In conclusion, the study highlights the need to use evidence-based practices in graduate education, as compared to the now common training of graduate students based almost exclusively on the PI’s personal experiences. It also encourages the identification of reward mechanisms to acknowledge postdocs for their informal but pivotal role in training PhD students.

Copyright © 2016-2020 STEMM Leadership


  1. Up until this semester, I never really thought of concept as such where a certain type of mentorship model makes an actual difference in the results. Outside of this article, I am currently working with Dr. Gutzler on her research about how the efficacy to a graduate assistant is affected by his or her mentor given Georgia State’s apprenticeship model. Reading this article reminded me of that because so far we have found that a mentee being able to not only see how a Graduate Assistant runs the classroom but also doing weekly checkins with their mentor as well as the professor in charge not only increases efficacy in the apprentice but also in the actual mentor. To elaborate, I believe that mentorship not only benefits the mentee but also the mentor because it allows for room to continue to grow and learn on both ends of the spectrum.

  2. This is a fascinating and very smart concept. When most of us have mentors, they are usually in different fields or if they are in the same field they have slightly different paths. I think if you have someone on the exact same educational path as you mentoring you then that would be a very productive relationship. One of my mentors is a Doctor which is great, but not my exact educational and career path since my goal is to be a Physician Assistant. It would be amazing if there was a program like this at whichever Physician Assistant Masters Program I decide to go to. It would be nice and encouraging to have someone who had seniority in the same program on the same path as me to help lead and guide me through the tough process. This would be immensely motivational for me on that 2 year journey to degree completion and certification.

  3. This article is one I can definitely agree with as it is something I have not heard of before, but I believe it sounds like a great idea as a someone who is interested in likely pursuing an MD-PhD such as myself. I am currently in a research lab with my mentor, who I am lucky to relate with, which aids in my ability to learn and understand the basis of his developmental research. I believe this lab has helped me see that I enjoy research and will likely pursue a degree in which I can continue doing research. However, it has also made me wonder what will it be like with a different mentor?
    I feel grateful I am able to learn from my mentor, but I know there are also mentors out there with different ideas that require different teachings that may not be as easy to understand, so what will that be like? I believe the idea to continue to explore the cascading mentorship approach would make learning new concepts in research less grand and a style that will help make the transition for a researcher into a lab more comfortable as they can likely relate more from their peers than they can from a mentor with years of experience.

    • Lola, you have made very excellent points. I am thrilled that you’re in a lab with a mentor that you can relate to, and I believe that it will aid in your journey to your MD-PhD tract. After reading the article, it also discussed “cascade mentoring” I think that will be beneficial because, as you’ve started, you want to experience what it would be like to have another mentor. With cascade mentoring, you will have the opportunity to be mentored by people who may relate better to you or have had more recent experience in the field you’re exploring. While I believe that senior mentors will be essential to your overall growth as a Ph.D. student, someone who is more engaged and recently experienced in what you’re currently experiencing could allow you to feel more secure in your future and current work. I would like to speak to you about your interest in research and how it has driven you to want to pursue an MD-PhD. In college, it is hard to find your niche, and I would love to have this same cascade mentoring with you to see how I can find something that drives me to find my niche.

  4. I agree with the cascading mentorship model. I had the previous notion that the cognitive apprenticeship model was how a mentor-mentee relationship had to function. I thought regardless if you’re a postdoc, Ph.D. student, master’s student, or undergraduate student, we all had the same mentor in the lab, the PI. In my experience, my PI mentor relationship is not as strong as my postdoc relationship, and this could be due to the PI having to pour their knowledge into many people. The cascading mentorship model is the most effective way to pass down knowledge from senior advisors down to an undergraduate. I work directly under a postdoc; he is much more approachable, and his knowledge and skillset are just as valuable to me as my PI. PI’s roles have shifted to one that involves less direct contact with students, as the article stated, and while my PI is not always in the lab, there are still postdocs or Ph.D. students who I can ask questions and receive advice. Even though I am still an undergraduate, I have found that younger undergraduates are coming to me for questions. I never considered that I, too, could pass on my experience to those entering the field, becoming a mentor myself, and I’m eager for the role.

  5. I like the idea of the cascading mentorship model. What better person to get advice from than someone who has already been through the process and knows the struggles of being a PhD student. Even though this mentoring only helps with the advancement of the student, I still think it is worth it. As a future med student I would hope that I’m able to receive this kind of guidance. I think receiving help from a mentor, people would being able to enhance and learn certain skills that are good for their career.

    • I hope you find a good mentor, too! I am nowhere close to a Ph.D. or even a Master’s degree, but even as an undergraduate student, I believe cascading mentorship is key to a student’s success. Getting a Ph.D. is a big commitment and a long process. For our present and future professionals, both good communications skills and professional skills are expected of a person at Ph.D level than they did before. So, even undergraduate students are now required to develop these skills because college degree is the new high school degree. I notice that colleges now are more willing to invest in people with passion than talents because talented people could give up when faced with unusual challenges, but people with passion will stick around. Best of luck to you on becoming a med student and pursuing your dreams!

  6. This is a very interesting idea and something I had never thought of in the past. I think most of us know that anyone can have a mentor, but what we may not know is that most of us are actually also qualified to serve as a mentor. As a Masters student I would not normally think that I would be in any position at all to mentor an undergraduate student, but this research suggests that that could be beneficial to me and to another student. College and grad school are places you go in order to learn the skills you may need in your future career. In science, so much of that is being able to work with and teach others about the work you are doing, in order to drive the progress of science itself. Being a mentor to a more junior member of your organization seems like a good way to practice this and gain valuable experience that can allow you to be more successful in the workforce early on in your career. This is something I would like to see more of in academic institutions, where others are actively seeking out individuals that they can help, and in turn help themselves.

    • You’ve made a really nice point here Nicholas. I too have never even considered myself to be a potential mentor in grad school, but after reading your statement and how you’ve explained it, it makes a lot of sense. Mentors don’t have to be a PhD or an MD, mentors can even be undergrads further along in the program. I think that a mentor doesn’t have to be someone with high titles, it has to be someone who has a lot of first-hand experiences. Experience, especially in the workforce, is extremely important and how can we as people grow and learn if we’ve never experienced the events that make us gain these valuable lessons and skills. We have valuable information for undergrads, we give them the do’s and don’ts with classes and professors, we tell them classes to avoid and we tell them this because it’s what we’ve already experienced and what they will soon experience on their own. Until now I wouldn’t consider my myself a potential mentor but recalling on my past experiences has made me realize that when I would run into past classmates they remembered what I told them and they’d tell me “I’m so glad that you told me to pay attention to this and study for that, if you hadn’t told me that then I don’t know how I would have passed that class.” The other thing your post makes me think about is that we have to be potential mentors for other colleagues, because even though they may have less experience than us or less time in school than us, that doesn’t mean that one day they might teach us a thing or two in the future. We must make sure that when we get older and advance further into our careers that we remain open minded and think about the people who are following a similar path as us who are gaining new experiences and skills along the way to share to future people. Being a mentor, even for a brief time, can cause a positive ripple throughout time that can benefit future generations.

  7. I believe universities should increase focus on the “cascading mentorship” model because it is a mutually beneficial relationship for the university, postdocs and senior graduate students and the research results provide concrete evidence of research skills growth. I also believe the cascading model allows more relatable relationships to be formed since many of the senior graduates aspire to become Postdocs and want to absorb the information and have an opportunity to be shown what is needed to be just another level up the ladder. What is really effective about the model is how it continues to work as you move down the ladder (senior graduate students mentoring junior graduate students) because each mentor can not only tell but show the mentee what it will take to reach the next level is not only limited to research skills but can also include sharing of needed competencies. As the postdoc works directly with the student on a regular basis in the lab showing them proper techniques and giving them invaluable advice, trust and a sense of familiarity strengthens the relationship. I think the trust encourages both parties to consistently give their best efforts and growth of research skills occurs for the postdoc and the senior graduate. This leads to more chances of groundbreaking research for the university which will enable the university to increase their accessibility to funding and research opportunities along with adding more prestige to their name.

  8. Though more research is in fact needed to fully understand the “cascading mentorship model”, in my experience, I agree with its claims. As a graduate researcher in a master’s program, I always seem to gravitate towards the PhD students in the lab when asking questions instead of asking my PI. This is not the only instance when I see this model take place. As a lab TA, I often notice my students asking my apprentice questions before they ask me and asking me questions about material from their lecture portion instead of their lecture professor. The article claims that “as the practice of science has shifted toward larger team enterprises and an increasing pace and volume of workload, the nature of the PI’s role has shifted to one that often entails less direct contact with students” but I would disagree with this statement as my PI has weekly meetings with us and constantly goes out of their way to be included in our research and our lives as a whole. I believe that the reason the cascading mentorship model occurs is because people feel more comfortable asking questions and being wrong in front of lower-stake people or people that are not grading/rating them directly. Instead, in my experiences, it is easier to be wrong in front of your peers instead of your boss and you are able to benefit more from not being concerned about the risk of being wrong as this allows you to ask all the questions that you have instead of holding back out of fear of looking less than.

  9. The “cascading mentorship” model, in my opinion, is the most useful model to share and develop skills inside a team. Univeristy lectures are more focused on the theory of chemical or biolgical processes. When it comes to practice, the know-how to perform those processes, accumulated by senior PhD students, is essential for younger ones in order to develop a more complete critical approach in their research activities.

    • Marco,
      I completely agree with you; University lectures typically focus on the big picture while senior Ph.D. students know how to perform chemical and biological analysis of the processes discussed by the University lecturer. Ph.D. students have valuable knowledge and possess the skills to teach it to undergraduates effectively despite not have formal training.

  10. As an undergraduate research assistant, I have personally experience the “cascading mentorship” model in action as well as witness many other graduates and undergraduate research assistants working and learning under this model. As the article has mentioned, this model truly benefits all most all members of the lab if it is utilized in the right way. By receiving direct contact mentorship from the more experienced postdocs, PhD, and master students, the less experienced graduate and undergraduate members of the lab are more likely to develop research skills faster. This model of mentorship reduces the stress and pressure on the less experienced members of the lab by creating a more casual learning environment as well as stimulate growth in research experiences. Moreover, junior students are often able to better connect with the senior students and postdocs due to the shorter age gap, which leads to a more “sibling” like relationship. In fact, this academic sibling relationship is highly regarded and treasured in many Asian cultures, especially in China, Korea, and Japan. However, as the article stated, this significant mentorship model does not receive enough recognition for its value in most Western cultures. Hopefully, as the STEM field becomes increasingly globalized, Western culture will also adapt the Asian mindset to better recognize the significances and values of the “cascading mentorship” model.

    • Thao,
      I agree with you that this model of mentorship reduces the stress and pressure on the less experienced members of the lab. I remember going into my first lab experience being afraid to ask questions around the PI when I first met him because I didn’t want to ask a question that may have seemed “ignorant” in his eyes. I was lucky enough to have been placed under someone doing bench work in the lab who welcomed any and all types of questions and was eager to mentor me throughout my project. The other graduate students in the lab were also that way and the stress and pressure on me was so much less. I also think that this “cascading mentorship” model helps in developing relationships in the lab as I am still in touch with those lab members today. I experience this same type of mentorship in the lab I am currently in an can attest to the fact that I have felt very comfortable asking questions and really getting learn the skills in the lab. It’s interesting to hear from you that this kind of relationship is highly regarded and treasured in many Asian cultures. Hopefully, this paradigm-shift towards the cascading mentorship model continues to persist and the development of PhD research skills continues to progress positively. Unfortunately, I know not all labs are this way and I think the issue is being able to have the right people in the lab that are less about themselves all the time and are more open to helping those around them.

Leave a Reply

Your email address will not be published. Required fields are marked *