Collateral damage

Nov 26 2012 Published by under Careers, Life

Over the Thanksgiving break (hope you all had a good one!) I found myself thinking about what happens when a PI commits fraud. This in part was spurred by the uptick in reported cases of fraud in science over the past few years, and in part by DrugMonkey's post on the RePAIR program for rehabilitating fraudsters. I have no idea whether this apparent uptick in fraud is due to more fraud occurring (probably) or simply more people being reported (also probably). But that's not what I was cogitating on. What I was wondering about was the fate of those directly effected by a PI committing fraud. The graduate students, postdocs and technicians in the lab. The various co-authors of publications that end up being retracted. What happens to them?

Let's make the assumption here that the PI is the sole perpetrator of fraud in a lab. One can always make the argument that the lab personnel should have been aware and/or that they should have been more diligent about checking the data that went into a publication that bears their names. I certainly preach to my own group that if their name is on a manuscript they share responsibility for what's in it. But let's face it, there are many ways a PI could fabricate data without raising people's suspicions. "Hey, look at these gels I ran while you were away at the conference last week." Despite what many would argue they should do, many junior peeps would be very reluctant to question the boss. And what if the fabrication occurs in grant proposals the lab peeps never get to see?

So what happens? Are all the lab personnel tainted? If the PI commits fraud in grant proposals and is caught, does that stigma stick to everyone in the lab? Would any of you PI's out there hire someone relatively fresh out of a lab where the PI had been found to be a fraudster? I honestly don't know what I would do were someone with that kind of tragic background to apply to join my lab. I'd like to think I'd try to be as objective as possible, but how do you weigh their accomplishments versus other applicants? What can you trust?

Even worse, one can easily imagine the careers of grad students and postdocs being absolutely destroyed when their first author publications are retracted due to the PI fabricating data. How would these folks compete in today's very competitive market when they've lost many, if not all, of their publications?

When a PI commits fraud we tend to focus on that person and what should happen to them. But instead of worrying about someone who destroyed their career by attempting to cheat, instead of trying to reform them, shouldn't we be trying to minimize the collateral damage?

We're focusing too much on the perp, too much on the funds wasted and fake science published. Don't get me wrong, that's all bad stuff. Let's just not forget about the more immediate victims.

22 responses so far

  • DrugMonkey says:

    Keep in mind that some subset of the trainees may also be complicit. So how to sift through who is tainted and who is not?

    There was a case profiled in Science awhile ago which tried to track the lab members in a case such as you propose. I seen to recall they had all been shelled out at hte follow up review. (dunno the chisquare, we could debate what the expected outcome would be for a half-dozen trainees....)

  • Bashir says:

    I know in some of the well known cases the grad students have essentially had to start over in another program. A 5th year grad student might scrap everything they've done and just go back to the drawing board in another program.

  • odyssey says:

    Keep in mind that some subset of the trainees may also be complicit. So how to sift through who is tainted and who is not?

    Right. Even in my hypothetical example where the PI is "solely" responsible, how can you tell?

    Bashir,
    How many start over versus just quitting science? Either way, assuming they're innocent of all fraud, it really sucks for them.

  • DJMH says:

    Honestly it is a bit difficult for me to imagine bench science cases where the PI is guilty and *all* the trainees are innocent as lambs. Like they're never going to run another sample after the PI finds one that works?

    I think the more plausible supporting scenario is in a more computer-based lab, where someone can fiddle the numbers of N or just wholesale outcomes, as with the psychologist a year or two ago.

    I would be pretty hesitant to take on a trainee who had been first author on a publication that was shown to be fraudulent, in my bench-oriented world. Hard to see how the PI pulls that over.

    • odyssey says:

      DJMH,
      A lot depends on how the lab functions. I can easily imagine bench science PI's being able to do this.

  • drugmonkey says:

    DJMH-

    Lot of p-shop band cutting going down in retracted papers and fraud cases....can't the PI be on the hook for that?

    • DJMH says:

      But are we talking about a PI running her own gel, and photochopping it? Or taking a grad student's ugly gel and chopping it? The grad student would necessarily know what was going on in Scenario 2.

      I just think as a grad student I'd be suspicious if the only data we had for a particular part of the paper came from the PI running some mysterious experiments. But as odyssey points out, this is going to be highly lab-specific...

      • DrugMonkey says:

        but the student would also know if the N was suspiciously different or if PI's "reanalysis" of the data set turned up something she didn't originally see....

        • DJMH says:

          I guess the fabrication being PI-only would be an easy case to make if all the problems were in grants, not papers. In some cases that may be a relevant distinction.

  • morningstar says:

    This is a timely post that hits to home as I was recently contacted by my former graduate PI that some of my publications from my time in the lab have to be retracted. I can say first hand that fraud absolutely can occur without the knowledge of the trainees (at least some if not all). In the truly crummy system that we have, co-authors can be asked to sign a retraction without knowledge of why the article is being retracted. As I have found out, you can talk to the PI, you can talk to officials at the university, but no one is under any obligation to tell you why the article is being retracted. In fact, the university is not permitted to discuss the matter if there is an open investigation. So what is a former trainee to do? Refuse to sign the retraction? In the end, I signed a vague retraction notice that indicates there may have been issues with the previous work I based my publication on, even if my work was completely on the up and up. I still don't know what the "issues" are. It is a very frustrating and disgusting process to be pulled into the middle of.

    • DrugMonkey says:

      I still don't know what the "issues" are.

      I don't see how this could possibly be fair to ask a co-author to sign a retraction without any explanation of why. For all this person knows, they could be about to be thrown under the bus by the PI!

  • bill says:

    Anecdata, slightly tangential: my last PI made a point of hiring a whistleblower from a lab where PI fraud was pretty clear but remained "unproven" due to the feeble ORI process.

    I don't mean to disparage those who don't blow the whistle, I understand why they make that choice, but I do think it's important to cushion the blow for those who can be clearly dissociated from the malfeasance. A systemic commitment to this might make whistleblowing a much less risky proposition, so that more innocent otherwise-bystanders would make that choice.

  • odyssey says:

    Okay, so maybe my hypothetical case of a PI-only fraud case is just a little improbable, and that perhaps one or more co-authors are involved in the fabrication in some way. Nonetheless, in these days of many-author papers, there are likely co-authors who are not guilty and will have their careers damaged or at the very least tainted. Rather than waste time and money on nonsense like RePAIR, shouldn't we as a community be trying to figure out whose those people are and try to help them? Or are we callous enough to just discard them because there are, after all, plenty of trainees coming out of honest labs?

  • PSG says:

    It was nice to have some perspective injected into the discussion, odyssey (re: the OP). I appreciate it.

    Where do the data come from? In the case that I am aware of, they came from the "collaborators" - and the collaborators were told they came from the lab. In reality, they came from the PI's computer, but hey, who wants to squabble over petty details, amirite?

    As for taking on those of us REALLY screwed by PI schnangans, remember just how tough these sorts of cases are to prove and just how easy it is to leave a lab as either a grad student or a postdoc. You don't expect that from your PI, but when you realize that it is happening, you have to document, document, document until you are sure that you have enough to make it stick - because who is going to believe you over your PI, srsly - then let the wheels of justice work ever so slowly. You still got to work - you still got to eat.

    I mean, it is one thing to leave a lab where the PI was suspended/found guilty, another where the PI was not. And yeah, best case then we're stuck hoping someone we know knows someone who is willing to take a chance on us, after we've lost time due to retractions/erratums and the insanity/chaos of an ORI investigation. Ruined careers? Ha. Let's just hope you did your networking...

    So high horse it all you want - they should have KNOWN - but some of you come off like a whole lotta graybeard doing it.

  • morningstar says:

    I don't see how this could possibly be fair to ask a co-author to sign a retraction without any explanation of why. For all this person knows, they could be about to be thrown under the bus by the PI!

    This is exactly why the system is so terrible. As a co-author it puts you in a very precarious position. However, as I have come to find, there are unfortunately few options. I have heard of co-authors refusing to sign retractions. What will that make individuals in the field think though? Is that better or worse for the innocent people involved?

    My university told me to sign the retraction but was quick to assure me there is no "guilt by association". However, I can tell you that is exactly how it feels being in that situation .

    • odyssey says:

      Looks to me like the institution mishandled the sexual harassment stuff (only probation? deleting the letter that should have been kept on file?), but as best I can tell did the right thing with the scientific fraud. My bet is it will have little, if any, lasting effect upon that place. Unless you're closely connected to Smart.