January 2007--Even in an institution that houses some of the world’s leading researchers and consistently produces breakthrough science, Julie Gottlieb waits for the other shoe to drop. That’s because, as assistant dean for the Office of Policy Coordination, Gottlieb must move front and center if research of the wrong kind surfaces, if there’s a hint of fabrication, falsification or plagiarism. No scientific institution is immune, she says, to the dark side of scientific enterprise.
Both motive and opportunity to taint science have increased over the past decade, whether in grant applications, research data or publications.
“Funding’s grown more competitive,” says Biological Chemistry Director Gerry Hart, with the School of Medicine’s Standing Committee on Discipline, bringing unusual urgency to produce a track record. “Plus,” he says, “we’ve had this pressure that borders on the absurd to publish in top journals.”
Throw in digital technology—it simplifies malfeasance—and the pieces are all there.
“At the very least, the public is more aware of research fraud than before,” says Stanford’s John Brauman, an editor at Science. “And that leads to the perception that it’s happening more frequently.”
Such perceptions have been stoked by high-profile misconduct, such as South Korean scientist Woo-Suk Hwang’s bogus human cloning and former Vermont professor Eric Poehlman’s fabricating research grant data on menopausal occurrences.
The Hwang incident in particular sparked action. Several groups have re-evaluated their scrutiny; Brauman headed a panel appointed by Science to recommend changes in the journal’s handling of papers. Likewise, the Council of Science Editors just completed a white paper on promoting integrity in scientific publishing. And while the government has long barred errant researchers from ever receiving federal funding, more punitive measures may be on the way; this past summer, Poehlman was sentenced to a year in prison—the first time academic misconduct resulted in incarceration.
But while many approve a tougher stance, others like Brauman remain cautious. One downside, he explains, is the effect on the relationship between scientists and the public: “It’s extremely important that the public has confidence in science and sees it as having the utmost integrity.” Stepping up measures to spot and punish offenders may flag nonscientists that misconduct is becoming more pervasive. Also, such broad tactics may not be necessary.
“Research is inherently self-correcting,” says Hart. “Unless errant scientists’ work is highly obscure, they’ll get caught.”
At Hopkins, instances that have arisen have exacted a high toll. Each case can take up to two years and thousands of man-hours from initial inquiry to final resolution, says Gottlieb: “But since we’re dealing with someone’s career, it’s both important and necessary.”
For all these reasons, the Office of Policy Coordination works to nip misconduct in the bud. While students with federal training grants are already required to attend an ethics course, the OPC hopes to implement a school-wide standard for misconduct education, particularly proper data management. And committee members are surveying faculty and students to learn what prompts misbehavior and thus plug potential holes.
Inadequacies in civility and communication are common threads. “What we often see in labs involved in misconduct allegations are poor interpersonal relations and less clarity in defining roles,” says Gottlieb. Trouble begins, for example, when PIs treat postdocs as technicians rather than trainees.
Sheila Garrity, an OPC assistant director, says scientists need to be more aware of the full scope of impropriety. “It’s not uncommon for scientists to massage visual data with Photoshop or publish replacement data to help expedite publication,” she says. “Many people don’t appreciate that this is fraud, every bit as serious as outright data fabrication.”
Even earlier, in preparing grant applications, investigators must toe the line, Gottlieb says. “We review allegations of suspect applications using the same standards as for journal articles.” In short, says Chi Dang, vice dean for research, “we don’t want our faculty to play with fire.”
The OPC plans to develop online materials, host guest lectures and publish in internal media to get the message out. They are also considering appointing “ambassadors” from faculty ranks to help communicate proper rules of conduct.
“In the end, it comes down to how we manage our labs,” Hart stresses. Simple practices can help: Hold regular lab meetings; check lab notebooks; and set an atmosphere that champions truth. Though playing lab policeman may stress a lab leader’s schedule, what’s lost time compared to sowing bad science and losing public trust?
Peering more closely at peer review