Research Theory is a nonprofit organization dedicated to fostering creativity and innovation in scientific research through the ‘science of scientists’. We do this by supporting research into management and mentorship practices in academia, and developing evidence-based resources to maximize scientific achievement. We believe academia needs a new theory for how we do research with intentional practices that support our collective goal of scientific progress.
Declining innovation amongst a crisis of well-being
Since World War II, management science has revolutionized private industry. Academic science has largely been ignored, leaving policies and practices at the level of individual laboratories, departments, universities, and federal funding agencies subject to arbitrary decision-making by individual stakeholders. Despite the undeniable impact of scientific progress, science is becoming increasingly inefficient: since the 1930s the effective number of researchers in the United States has grown by about 4% per year, but research productivity in terms of societal gains is decreasing by about 5% per year across scientific fields.
There are many possible mechanisms underlying this productivity gap. One major area of concern is academic culture. A 2020 survey by the Wellcome Trust found that only 57% of researchers agreed that the research culture in their working environment supported their ability to do good research. In that group, only 65% of researchers believed that high standards and integrity were valued within the research community, and 75% of researchers agreed that creativity is stifled due to an emphasis on impact. Together these statistics suggest that the current incentive structure for research has a detrimental effect on scientific outcomes.
Source: What Researchers Think About the Culture They Work In, Wellcome Trust and Shift Learning, 2020
Behind these large-scale cultural trends is an undeniable crisis of well-being and career longevity in academia. A study published in Nature Biotechnology in 2018 surveying 234 institutions found that 41% of graduate students had anxiety and 39% had moderate or severe depression. A survey commissioned by the journal Nature in 2020 found that 1 in 4 postdoctoral fellows reported experiencing harassment or discrimination in their workplace, and that 51% of postdocs considered leaving academia because of anxiety or depression related to their work. In 2022, both Science and Nature published editorials raising the alarm on professors being unable to hire postdoctoral fellows, citing that three-quarter of survey respondents had challenges recruiting last year, with job postings left unfulfilled for months or even a year. Put simply, the system today is scientifically inefficient and personally unsustainable.
The mentor-mentee partnership as a unit of innovation
While the impacts of science are larger than life, the day-to-day process of science is deeply human. In most academic research groups, graduate students and postdoctoral fellows work alone or in small teams, striving to make progress on a research project in hopes of eventually publishing their work in a peer-reviewed journal. Professors in life and physical sciences are rarely at the theoretical or experimental ‘bench’ themselves; rather they guide and support the members of their research group who execute the day-to-day work of scientific research. This relationship is typically considered that of a mentor-mentee akin to a traditional apprenticeship, but in reality it is a two-way partnership. Because mentorship is the major mechanism through which professors engage directly with the process of science, mentors rely on their mentees for ideas, technical expertise, and ultimately the body of work that will be published. While the word ‘mentorship’ accurately implies that the relationship is critical for training the next generation of science, it overlooks how equally essential it is for producing the science of today. Given the structural nature of academic work environments, the two-way partnership between mentors and mentees is the greatest tool we have for scientific progress.
Intuitively, the mentor-mentee relationship matters a great deal for professional and personal success of both parties. Scientific achievement by the mentor strongly predicts mentee success along multiple parameters: for example, Nobel prize-winning mentors are more likely to have Nobel prize-winning mentees. But causality is notoriously difficult to untangle, as successful professors tend to recruit talented mentees, who might succeed regardless of the training environment. One study led by Dr. Brian Uzzi, professor at the Kellogg School of Management at Northwestern University and co-director of the Northwestern Institute on Complex Systems (NICO), found that mentees who graduated 10 years before their mentor won a scientific prize are 5.7x more likely to win a prize of their own, 4.3x as likely to be elected to the National Academy of Science, and on average have double the h-index as peers in similar training environments but whose mentor did not go on to win a prize. Did mentors with prizewinning capacity transfer some tacit instruction that supported their mentees’ ability to win prizes of their own? Or is something else at play? These studies demonstrate the potential significance of mentorship in scientific achievement, but also reveal a major challenge: a lack of direct evidence of even good proxies for evaluating mentorship.
Feedback and accountability for mentorship
Direct evidence and good proxies do not exist, because data on mentorship in academia is almost totally lacking both internally and publicly. Professors are largely expected to learn how to mentor on their own: the 2020 Wellcome survey found that of almost 2,000 researchers in managerial roles, less than half had received any training on managing people. That same survey found that implementation of management practices is shockingly poor. Graduate students and postdoctoral fellows were asked which behaviors typically associated with effective management they experienced in the last year–on average they responded ‘yes’ to only 4 out of 14 (see chart below for a full breakdown). At odds with these poor outcomes are professors’ confidence in their own abilities: 80% of faculty believe they have the knowledge and skills to manage a diverse team and support others with their professional development, indicating that a lack of feedback is a critical breakpoint in the system. Beyond feedback, there is a significant lack of accountability around mentorship–only 44% of mentors agreed that good management and leadership is recognized at their institution.
Source: What Researchers Think About the Culture They Work In, Wellcome Trust and Shift Learning, 2020
In practice, mentorship is extremely variable. Depending on the personalities of the individuals involved and size of the research group, mentees can experience anything from total autonomy to extreme micromanagement from their mentor. While data are lacking, researchers agree that short-term thinking, intense pressure, and toxic power dynamics are tolerated as the status quo.
Over half of researchers surveyed report witnessing or experiencing bullying or harassment. Less than half agreed that they would feel comfortable reporting instances of compromised research practices without fear of personal consequences, and 23% of junior researchers agreed that they had been pressured by their mentor to produce a particular result. These statistics are the tip of the iceberg, but illustrate the consequences that a lack of coherent standards and expectations for scientific mentorship can have on both the scientists themselves and the research they produce.
We argue that the dual failures of feedback and accountability for mentorship are directly related to a slowing pace of innovation in the United States. Ad hoc management carried out by unqualified mentors has fueled a culture that prioritizes short-term gains by whatever means necessary, be it harassment, bullying, or shoddy research. This stifles creativity, slows innovation, and undermines the quality of work produced.
Our strategy to develop a new theory for research
Research Theory’s goal is to change the culture around mentorship in academia in order to support creativity and innovation. To start, we will pursue the following areas:
Collect data on the state of mentorship and management in academic scientific laboratories. This is a highly neglected research topic that will serve as the baseline for future studies and evidence-informed interventions.
Invent metrics and tools for measuring mentorship that are dynamic, scalable, and reflect the two-way nature of a scientific partnership. While the mentor-mentee relationship is complex, simple quantifiable metrics are essential for accountability and tracking progress.
Recenter creativity at the heart of the scientific process. While academia has much to learn from industry management practices, it has its unique challenges. What can laboratories learn from how both factories and art studios are run?
Develop resources and interventions that implement research theory practices. This includes training for faculty members, team-based training for the research group, and resources for graduate students and postdoctoral fellows as they navigate their research and professional environments.
We are building a coalition of researchers, academic institutions, and funders to develop a new theory for how we do research. Join the effort by emailing us at info@researchtheory.org.
Elaine Sevier is President and co-founder of Research Theory