How To Support Graduate Education in the Sciences?

Frederick Grinnell is Professor of Cell Biology and founder of the Program in Ethics in Science and Medicine at the University of Texas Southwestern Medical Center, Dallas. His newest book, Everyday Practice of Science: Where Intuition and Passion Meet Objectivity and Logic offers an insider’s view of real-life scientific practice. Grinnell demystifies the textbook model of a linear “scientific method,” suggesting instead a contextual understanding of science. Scientists do not work in objective isolation, he argues, but are motivated by interest and passions.  In the article below he looks at how graduate scientific education should be changed.  Read previous posts by Grinnell here.

In the 31 July issue of Science, Jeffrey Mervis’ News Focus discusses funding of graduate education in the sciences. Currently, the overwhelming majority of U.S. graduate students are supported by research grants to faculty members who hire the students to work in their laboratories. Some research leaders argue that the U.S. would benefit if more graduate students were supported by independent fellowships instead of research grants. For instance, independent funding would empower young investigators to follow their passions, be more creative, and function less as technical assistants – “a pair of hands.” However, others are less enthusiastic about the wisdom of removing from individual investigators the responsibility for selection of students.

Primarily, Mervis’ report concerns the economics of graduate education and the roles of the graduate student workforce in advancing university teaching and research interests. The article fails to address an ethical dilemma of the current system – the inherent conflict that occurs regarding mentorship. When graduate students are supported by research grants rather than by training programs, they become laboratory employees as well as trainees. Laboratory directors have a fiduciary responsibility to their grants to ensure that employees carry out the proposed research and maximize productivity. At the same time, investigators have a mentorship responsibility to ensure that trainees receive the education and experience necessary for future success.

What is in the best interests of laboratory productivity might not be in the best interests of a trainee. It is not enough that graduate students learn the technical skills – “a pair of hands” — to do experiments. Just as important is that they learn how to design experiments. Good mentoring may mean allowing students to design experiments even if doing so increases the likelihood of failure, an outcome that certainly does not advance laboratory productivity. Similarly, allowing students to write the first draft of papers may slow down the total time it takes to publish the research. Nevertheless, learning how to write papers is a key skill for students to acquire. Finally, imagine the situation of an advanced graduate student who comes up with a novel idea that she would like to explore. From the point of view of training, perhaps this development should be encouraged. But if the laboratory director is trying to finish up a paper for an upcoming grant submission, then the student’s tangential efforts likely will be discouraged. It is not the right time.

In addition to teaching stipends, three types of mechanisms commonly are used to support graduate students in the sciences: research grants to individual investigators, department/program research training grants and independent student fellowships. Amongst these three mechanisms, perhaps the time has come to transition to training grants as the primary mechanism of support. Training grants retain local institutional control over graduate education and empower local programs (vs. individual laboratories) to define more clearly what education should entail and how long it should take. More robust consideration becomes possible about how graduate science education can advance career goals beyond the traditional path towards academic scientist. Students gain flexibility. Nationally, strategic planning for graduate education will be facilitated. Deciding how much money to allocate for research training grants can influence the size of the graduate student workforce. Deciding which institutions receive training grants can nuance multiple outcomes such as best scientific opportunities vs. other important factors, e.g., the marginal impact of training grants on traditionally underfunded institutions and the communities in which they are located.

Thirty years ago, one and two author research papers were common; papers with many authors were rare. Now, research groups have become larger and collaborations more frequent. Papers with five to ten authors are common. Increased size and complexity challenge implementation of the traditional dual roles of mentor and laboratory director. Funding graduate students in the sciences through training programs rather than research grants will help focus attention on the importance of both these roles.