Read our COVID-19 research and news.

Two postdocs at the University of California, San Francisco, make a sketch of their career paths as part of the National Institutes of Health–funded Broadening Experiences in Scientific Training program.

Elizabeth Silva/2015 MIND program at UCSF

Making grad school work for STEM students

How many reports does it take to change U.S. graduate education?

Answer: anywhere from one to 20. But the current system must want to change.

The training of graduate students in science is no laughing matter. But the cascade of reports issued on the topic over the past quarter-century has become something of an inside joke among those who care about graduate education in science, technology, engineering, and math (STEM) fields. So, when a committee of the National Academies of Sciences, Engineering, and Medicine (NASEM) this week issued a report on “revitalizing” graduate STEM education that referenced 19 related studies, its chairperson wasn’t surprised.

“The first thing people ask me is, ‘So what’s new?’” says Alan Leshner, CEO emeritus of AAAS in Washington, D.C., (which publishes ScienceInsider). His answer, in a phrase, is the call for a “student-centered” education.

“The current system works well for the PIs [principal investigators], institutions, and federal agencies that get relatively cheap labor and churn out lots of papers in top journals,” Leshner asserts. “But it doesn’t work well for students, and for many employers.”

What’s needed, according to the report, is a greater focus on mentoring and high-quality teaching. Doing so, it suggests, would also address many other chronic problems—including inadequate preparation for careers outside academia, a lack of diversity, and overspecialization—that previous reports have cited.

A job too far

The misalignment between the career path that many faculty expect their students to follow, and where those students actually wind up, is a major theme of the new report, which is an update of a 1995 NASEM report on “reshaping” graduate education. The tellingly similar titles suggest the problem is long-standing.

Most professors erroneously assume their job “is to produce little clones of themselves,” Leshner says. Yet fewer than 40% of the country’s STEM Ph.D.s actually work in academia, the report notes, and fewer than half of that group become independent investigators like their mentors.

That mismatch has several unintended consequences, the report notes. It can push some students into unwanted postdoctoral positions as they sort out career options and cause others to abandon science altogether.

But although the problem is serious, Leshner says any changes should be “evolutionary, not revolutionary.” Students still need to learn how to design and carry out research and to communicate the results to a broad audience. A combination of workshops, internships, and networking opportunities that expose students to career opportunities outside the lab, he adds, should “supplement, not supplant” the time needed to acquire what the report calls “core competencies.”

Panelist Suzanne Ortega, president of the Washington, D.C.–based Council of Graduate Schools (CGS), explains what that means in practice. Academic tenure and promotion decisions now rely largely on research productivity, in particular, a faculty member’s ability to garner competitive grants to fund research that will generate publications in prestigious journals. “I’ve been on a lot of those committees,” says Ortega, a sociologist and former provost and graduate dean at several universities, “and I’ve never seen one that asked about innovations in mentoring or the quality of that mentoring as measured by students’ satisfaction.”

A road map for change

Leshner’s big idea is that federal research agencies such as the National Science Foundation (NSF) and the National Institutes of Health (NIH) should require grant applicants to provide evidence of quality teaching and mentoring in their proposals. The agencies should then use that information, he says, to determine who gets the money.

It’s what Sally Rockey, president of the Foundation for Food and Agriculture Research in Washington, D.C., and former head of NIH extramural research, calls “social engineering through the pocketbook.” In 2013 Rockey helped NIH launch its Broadening Experiences in Scientific Training (BEST) program, which made 17 awards to improve faculty mentoring skills and professional development training for students.

A former top official at both NSF and NIH, Leshner says such culture change won’t happen overnight. “We’re not totally naïve,” he says. Professional societies and higher education organizations also need to do their part, he says, citing efforts already underway by CGS and the Association of American Universities, which is also examining ways to improve undergraduate STEM education.

Change is possible if there is institutional buy-in, says panelist Keith Yamamoto, vice chancellor for science policy and strategy at the University of California, San Francisco (UCSF), which received one of NIH’s BEST awards. (NIH made only two rounds of 5-year awards and never intended BEST to be an ongoing initiative, but UCSF officials have pledged to continue the program after its NIH grant ends this year.)

Panel members say the increased emphasis on quality mentoring need not become another burden on faculty already struggling to meet the growing demands on their time. Leshner says it can even become a recruiting tool. “Schools like UCSF have shown that it’s doable,” he says, “and students might want to take note of institutions that have such programs when they are deciding where to go for graduate school.”

However, the overall impact on research productivity is less clear. “The message seems to be, ‘Keep doing everything you’re doing, but find ways to do it better,’” says a congressional staffer who was briefed on the report but is not authorized to speak on the record. “I don’t see how that is possible unless there’s less emphasis put on the amount of research being done.”

Yamamoto doesn’t agree. He says the key to a student-centered education is not for professors to do less research, but to do it in ways that focus on teaching students the skills and knowledge required to become a practicing scientist.

“Right now we have a flawed set of metrics,” he explains. “We place a priority on being first author in a prestigious journal. If we get rid of those metrics, we can move closer to the core competencies that students need to graduate. And remember, the goal is not to become famous, it’s to discover new knowledge.”