Sins of the principal investigator

As a columnist who uses my own limited experience to make sweeping generalizations about science careers, I’m sometimes approached by trainees in one pickle or another. This past winter, for example, I received such an email from a former student at a large research university in the United States. I don’t want to disclose her identity, for reasons I’ll explain shortly, so let’s just say that she’s an adult human, and her email address has the “@” symbol in it.

During the first year of her Ph.D. program, the student faced a standard dilemma: deciding which lab to join. It’s a choice that requires meticulous research, usually conducted by asking the third- and fourth-year grad students, “So, which advisers don’t make you work on the weekends?”

Yes, this particular tale of woe is veering into adviser-trainee conflict territory. You have my permission to pre-cringe.

There’s no shortage of stories of animosity between principal investigators (PIs) and students. Any student who claims to have an amicable, uncomplicated relationship with their PI is either lying or is someone whose lab you should join immediately.

When she found out about the investigations, shortly after joining the lab, she nonetheless decided to stay because, as is often the case, she thought that switching labs would be the greater of two evils. But as the years passed, she says, she felt like she had to work around her adviser’s ongoing drama on a regular basis, seeking help from collaborators when necessary.

Then, when she started looking for a postdoc, she faced additional unanticipated hurdles, including interview questions about her PI’s scandal. One potential employer even refused to commit to a decision until receiving reference letters from all six of her collaborators. And, she told me, at least one of her lab mates also interviewing for postdocs was asked about the negligence as well.

Uncomfortable questions and increased due diligence may sound like innocuous penalties. But this student felt like she was being punished for someone else’s negligence, branded as an outcast worthy of instant suspicion. Should she, she asked me, be punished for the sins of her PI?

My initial response was of course she shouldn’t! She’s no more responsible for his past misdeeds than the undergrad playing Candy Crush at the lab bench is responsible for awarding tenure.

But my experience with academic misconduct is limited to running potential plagiarists through the ringer of, so I posed this scenario to some actual experts: Adam Marcus and Ivan Oransky, co-founders of Retraction Watch, a blog reporting on mistakes and misconduct in the scientific community. It’s probably the most scandalous scientific meta-publication you’ll find—titles of just the most recent few pages of posts include the words “stole,” “confess,” “lawsuit,” “investigation,” “misleading,” “phony,” “plagiarism,” “fabrication,” “contamination,” and “under fire.”

Though they both felt that this student’s situation was a singular case, Marcus and Oransky agreed with my gut feeling that she shouldn’t suffer for errors that preceded her. After all, when the negligence occurred she was still in high school, where the only mistakes she was probably making were the ones high schoolers make, such as playing the clarinet.

It would have been a different situation, Marcus wrote in an email to me, if she had witnessed the negligence and chosen not to interfere. “That’s another matter entirely and one about which I’d be concerned as an employer,” he wrote, “just as I’d be concerned if I were, say, hiring at a hedge fund and a résumé came across my desk from someone who’d been at Bernard L. Madoff Investment Securities.”

But, Oransky pointed out, the fact that the student didn’t know about her PI’s history highlights a concerning issue about how institutions, and the scientific community as a whole, handle potential misconduct. “It is routine for U.S. universities to refuse to disclose reports of their investigations,” Oransky wrote to me. Getting your hands on these reports “often requires [Freedom of Information Act] requests, and even then you may not get much.”

In other words, science has no equivalent of sex offender registry laws. A PI with a tainted reputation is not forced to wear a scarlet “p > 0.05” upon his or her bosom, so new students may have no idea what they’re getting into.

Whether this should be the case can be a thorny issue. Although increased transparency could help trainees make informed decisions, for instance, some argue that it would also mean excessively publicizing accusations, even when the PI is ultimately exonerated. After all, saying, “It’s OK. I was only accused of misconduct” doesn’t exactly make potential grad students want to hitch their wagons to a PI’s star.

Perhaps for this reason, university officials sometimes believe it’s in their best interest to sweep scandals under the quad sod. I don’t know whether that happened in this student’s case, but according to Oransky, it has certainly happened at other schools. In the long run, though, he says, such cover-ups end up creating more problems for universities than they solve.

Transparency becomes even more important, not just to alert students to potential PI poison, but also to set the record straight about which students had nothing to do with particular instances of misconduct. “In general it would be a good idea for universities to release full reports of their investigations,” Oransky wrote, “rather than to, well, force innocent bystanders to prove their innocence.”

Granted, in the case of the student who contacted me, some of the responsibility may fall on her, which she readily admits. Rather than Google-stalk her prospective PI before joining his lab, she assumed that tenure meant everything was peachy—and then watched in horror as her lab experience became more complicated than what she thought she had signed up for. But she also believes that “every PI with pending misconduct investigations should have to disclose that information to a potential graduate student or postdoc.”

I would take it a step further. Every PI should also disclose data on the average time to graduation from her or his lab, the lab’s financial picture, and personal or political dynamics that could negatively affect the student’s progress. First-year students choosing their labs shouldn’t have to rely on rumors from older grad students, only to find out—oops!—we didn’t want to tell you this, but the PI’s ex-spouse is now the department head and will ensure you never graduate, our grants all ran out, and the whole lab is moving to central Australia next year. Who wants venomous snakes?

Science thrives on transparency. When we draw conclusions in the lab, we use all of the data points, not just the ones that line up. And scientists need transparency along our career paths as well. We’re responsible for conducting our own due diligence, but the onus also falls on institutions and those in positions of power not to obfuscate the negative.

All of which is a roundabout way of asking, “Do I have to work on the weekends?”

University of Hawaii lab explosion caused by inappropriate gauge

Elsewhere in Science: Pirating papers, handling harassment, and more