90

In a recent paper, a team of colleagues published results that are clearly made up. I am 100 % sure of that because my position at the university involves the management of the animals they used in the study, which grants me access to the native data.

Actually, the statistics were so poorly handled you have hints something is definitely wrong just by reading the abstract. My name doesn't appear on the paper, but I could be indirectly associated with that team through my position.

So how to properly react? Close my eyes and move on? Write to the journal they published in?

  • 3
    Are you saying that they're outright fabricating data, or are you saying that the published statistics do not support their conclusions? – David Apr 17 '17 at 20:28
  • 15
    Clearly used data made up out of the blue just to get significant results, but even though you can tell statistics are screwed up. Example: out of 3 main parameters compared between 2 groups of 15 individuals through several months, each single parameter had a similar (and very small) standard deviation in both groups, which is totally impossible considering the fact these particular parameters are quite highly variable within a single individual through time – Cobactan Apr 17 '17 at 20:34
  • 9
    Does your university have a code of ethics and/or an honor code that would allow you to report them without fear? – Michael Apr 17 '17 at 21:53
  • 13
    Is this in academic/government/ or commercial setting? (in research-active university setting in US, there will usually be a process that is pretty detailed and is designed just for this situation - that most faculty/staff probably won't know about - (except the faculty who've served on the various research ethics committees for some prior complaint). It is usually initiated by someone privately bring a concern to the research ethics officer (typically the Dean of Research) who then carries though any initial inquiry deciding whether to convene the next stages of the process. – Carol Apr 17 '17 at 21:56
  • 3
    Consider contacting your university's omsbudperson (if such a position exists) -- if it's anything like my university, they should be able to confidentially advise you on what to do/who to report this too. Additionally, what country is this in? We may be able to tailor our advice (e.g., by stating which government agency to report to if the work was publicly funded). – tonysdg Apr 17 '17 at 22:21
70

my position at the university involves the management of the animals they used in the study

If your immediate supervisor is to be trusted, go to him or her. If not, pick someone higher up in the university hierarchy. You are looking for someone who will

  • protect your identity as a whistleblower

  • investigate discreetly

  • have enough influence within the university to be able to trigger effective action

Edit (in response to some comments): I should really have suggested that you look up your university's policy and procedures. Let's look at what MIT has to say about "Research Misconduct" as an example.

Unethical behavior in research and scholarship strikes at the heart of the scholarly and educational enterprise. [...] Supervisors must enforce the highest standards for conducting research and creating and maintaining records of the research [...] specifically, laboratory and center directors....

Definitions: [...] Fabrication is making up data or results and recording or reporting them.

Duty To Report: Each member of the MIT community has a responsibility to report any conduct that he or she believes in good faith to be Research Misconduct at MIT. Ordinarily it is appropriate in the first instance for a Complainant to report his or her concerns to the supervisor of the prospective Respondent. [...] Consultation and guidance is always available from the Office of the Vice President for Research or from senior academic officers (deans, department heads, laboratory directors)....

A supervisor who becomes aware of possible Research Misconduct, either from his or her own observations or because of reports, has a responsibility to bring allegations of Research Misconduct directly to the Vice President for Research in order to ensure that proper procedures are followed.

  • 13
    If there is an established regulation/code of ethics, shouldn't the OP follow this instead of speaking to someone in his/her hierarchy? (Unless of course the latter can really be trusted on a personal level. But then again his/her role would be merely supporting in case the code exists.) – Helen Apr 18 '17 at 9:02
  • 13
    There is typically a designated person, often with a title like "ombudsman", for discussing such matters with. – chrylis Apr 18 '17 at 9:55
  • 1
    I feel that this should also simultaneously be reported to the journal. And potentially to the general public. Both institutions and journals have been known to, ah, ignore evidence of misconduct. The scientific public should know about this, and they should know about it as soon as possible (without neglecting due diligence of course). – Konrad Rudolph Apr 19 '17 at 14:09
  • 1
    @KonradRudolph, the institution may have a procedure for notifying publishers, so surely this path is favourable. (Of course, if the procedure is disregarded, then notifying the journal directly is the next step.) – user2768 Apr 19 '17 at 14:31
19

I believe that you are doing the right thing by not letting this pass - there needs to be a view of 'protecting' the integrity of science.

Most importantly, you need to protect your career etc. To add to @aparente001's excellent advice, I would strongly suggest that you make sure you maintain documentation and evidence. It will be absolutely vital for you to keep a copy of the documentation for any investigation (and keep a backup copy for yourself).

Having said that, before you take any action, you should once again, check and double check that your claims are valid. Your checks and proofs could form part of your evidence.

Having said all that, your name and position is indirectly associated to the research group, so once you have collated your evidence and discreetly reported this to the appropriate people, you realistically can do no more discreetly, as you've done all you can, unless you wish to approach the journal directly or go public.

  • “You can do no more” — Uh, yes, you can and should. You can inform the public. Institutes are wary of sullying their name even by association, and even if they do disclose the misconduct, they usually do so very late. – Konrad Rudolph Apr 19 '17 at 14:13
  • I was referring to going down the discreet route - looking at it - this was not clear - have added that point. – user70612 Apr 19 '17 at 14:18
  • Ah yes, that changes things. That said, depending on what you mean by discreetly, OP has the possibility of anonymously blowing the whistle, e.g. on PubPeer. This is only discreet insofar as it concerns OP, of course. For the other parties involved, it’s rather the opposite. – Konrad Rudolph Apr 19 '17 at 14:19
  • This could be a basis of a pretty decent answer – user70612 Apr 19 '17 at 14:20
11

Follow your University's Reporting Protocol

Most Universities, as well as most businesses, government agencies, non-profit organizations, and organized religions, have established protocols to help you through this reporting process and, generally, to ensure some combination of proper and timely handling of the issue, image protection, protection against attempts at retaliation towards whistleblowers, and protection against people being accused of things without basis.

For example, Humboldt State University provides the following relevant policy (source):

The Dean of Research, Economic and Community Development will serve as the RIO who will have primary responsibility for implementation of the institution’s policies and procedures on research misconduct.

and later it's specified that:

All University members will report observed or apparent research misconduct to the RIO. If an individual is unsure whether an incident falls within the definition of research misconduct, he or she may meet with or contact the RIO to discuss the suspected research misconduct informally, which may include discussing it anonymously and/or hypothetically. If the circumstances described by the individual do not meet the definition of research misconduct, the RIO will refer the individual or allegation to other offices or officials with responsibility for resolving the problem."

(RIO = Research Integrity Officer.)

Your University likely has a similar policy in place specifically for situations like this, and following the procedure in your University's policy is the best way to ensure your complaints are taken seriously.

  • The link you provided is the Policy and Procedure for Responding to Allegations of Research Misconduct. I tried to find the policy and procedure for reporting an allegation, and couldn't. If you know this university better, perhaps you could find it? – aparente001 Apr 18 '17 at 19:54
  • 2
    Thanks, that helped me find it. "Responsibility to Report Misconduct: All University members will report observed or apparent research misconduct to the RIO. If an individual is unsure whether an incident falls within the definition of research misconduct, he or she may meet with or contact the RIO to discuss the suspected research misconduct informally, which may include discussing it anonymously and/or hypothetically. If the circumstances described by the individual do not meet the definition of research misconduct, the RIO will refer the individual or ... – aparente001 Apr 18 '17 at 20:56
  • allegation to other offices or officials with responsibility for resolving the problem." (RIO = Research Integrity Officer.) I like your idea of adding this to your answer. – aparente001 Apr 18 '17 at 20:56
1

Are your colleagues federally funded? If yes, sue them on behalf of the federal government Qui Tam. Other nations may have similar laws.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.