Last week, Marc Tessier-Lavigne announced that he will resign as president of Stanford over work performed many years ago, in labs at three different institutions. While most of the attention has been focused on the fall from grace of this distinguished scientist, this sad situation carries broader lessons about avoidable outcomes.
A delicately worded sentence in the investigative panel’s report notes that there “may have been” opportunities to “improve laboratory oversight and management” in the Tessier-Lavigne lab. This understated conclusion rests on three elements identified by the panel: multiple people in the lab manipulated data over time, there were “oversights” in correcting the scientific record once discrepancies were called to Tessier-Levigne’s attention, and the lab culture was wanting in key ways.
While only four paragraphs in the full 95-page report are devoted to lab culture (more than 60 are focused on images that were manipulated), those paragraphs flag important issues that more of us who work in universities should heed. These issues are central to training the next generation of scientists, for leaders, and for the institutions that employ and educate scientists.
The report describes a lab culture that evolved over time as Tessier-Lavigne became busier and busier. At first, mentoring for scientific excellence and rigor was paramount. But eventually the lab evolved into a place in which members were apparently divided into winners and losers, based on their experimental results.
For some years, leaders of the scientific community have been concerned about the effects of hypercompetition on careers and the research ecosystem. In a parallel development, evidence has been accumulating about the importance of social systems and workplace environments — interpersonal climates — for the rigor and integrity of research. Research ethics is often thought about strictly in terms of personal misconduct, and not often enough in terms of interactions and structures within lab communities that tolerate, and sometimes implicitly promote, corner-cutting and rushing toward publication.
Think about it this way: If approval and rewards at work are directly linked only to results, not process, where do you put your efforts? When only desired data mean success, and when not producing “the right” data makes you a “loser,” the already fierce pressures are intensified. Or, as the scientific panel examining the Tessier-Lavigne papers phrased it:
“Perhaps science in general could benefit from a deeper recognition by senior scientists that they need to be mindful to defuse any putative pressure to please a Principal Investigator beyond the significant, ordinary pressure that postdocs already experience in a competitive scientific field.”
Add into this environment a rarely-available principal investigator who is juggling daily lab with other major responsibilities. Science editor-in-chief H. Holden Thorp’s recent observations about what it takes to lead a large lab is one important piece of this puzzle: Everyday questions get stored up for minutes of precious time that are only intermittently available. In that context, putting your best foot forward with “good” data is the direct route to approbation and success. From the panel’s report:
“The Panel does not find, based on the evidence available to it, that Dr. Tessier-Lavigne knew about the lack of rigor in certain aspects of the scientific process in his laboratory leading up to the publication of Nature ’09 or about the lack of quality in certain of the day-to-day scientific work and figure presentation for the paper.”
When I interact with early career scientists about ethical issues they face, they are not focused on the compliance-based topics covered in their required courses on the responsible conduct of research. Instead, they are struggling with more nuanced, real-world issues: Where is the line between making my data look better (prettying up) and unacceptable alteration? Who do I talk to about this when this pressure to “clean up” my images comes from colleagues or supervisors? What should I be doing when lab mates are belittled or reduced to tears in lab meetings? How do I advocate around authorship when my principal investigator minimizes my work or adds someone to a paper who didn’t contribute because “they need the paper”?
My colleagues and I at the National Center for Principled Leadership & Research Ethics believe that we must step back and reinforce our espoused values in our actions. Our premise is that research excellence is about more than what is achieved; it also encompasses how work is done, and by whom. This means that exceptional scholarship, done ethically, happens in inclusive working environments, with positive interpersonal dynamics, and in ways that develop, mentor, and support lab members to build their own productive, meaningful careers. Research groups and academic units do not only produce data and results: they also reproduce and sustain themselves — thus shaping future environments in the field.
We must incorporate into the formal and informal curricula skills for grappling with the everyday ethical issues faced in research. We see a need for developing leadership skills for individuals — who first lead themselves and then others — to build practical tools for teamwork rooted in personal values and goals. We promote in our writings and in our workshops the importance of open channels of communication in which, among other things, research group members can talk about pressures they are feeling, and the tacit value and reward systems that they feel accountable to — which may belie the more overt, public pronouncements.
We also believe that you can and should measure lab working cultures. For many years, working with Brian Martinson and Carol Thrush, we have been measuring the research integrity climates of university departments. With support from the Sloan Foundation and in collaboration with the American Geophysical Union, we have extended those measurements to include interpersonal elements, including how power is (ab)used and the civility of research environments. These unique, validated instruments, administered online, take a snapshot of department and lab cultures, measured by perceptions of working climate by those within them. The results allow institutions to find the bright spots where things are working well, to share those practices more broadly, as well as to discover environments where improvements are possible — or where trainees could use some extra support and attention. A key component is generating data about “early warning signs” that need attention before an actual misconduct crisis erupts.
Lab culture matters. If we truly care about the education and mentoring of those within our institutions and research environments, we should teach skills to support good interactions, and provide realistic recourse when things are not going well. Institutional and lab leaders should assess lab and departmental cultures, and they should use the resulting evidence to shape actions and policies. Attention to, and accountability for, these elements of lab culture is just as crucial to their role as generating valid “results.” Indeed, those two dimensions are inseparable.
C. K. Gunsalus is the director of NCPRE, professor emerita of business, and research professor at the Coordinated Science Laboratory in the Grainger College of Engineering at the University of Illinois Urbana Champaign.