scientific ethics dilemna (longer than I'd intended)

[Amber]

Without going into too many details, a lab I once worked in had a rather controversial paper published in one of the top three scientific journals. This paper was controversial because it showed "proof" of an interaction not thought possible wherein one certain reactant, acting on another, resulted in a loss of a product. Other labs had been unable to find proof, and in fact, had results showing that rather than the experiment providing a loss of the product, they resulted in a gain.

While recently talking to friends still in that lab, I found that upon repetition by other researchers in the lab other than the one who did the initial work which the paper was based on, the published results showing the decrease could not only not be replicated, but an increase was found. This repetition was done by three different people a total of eleven times, each time consistenly showing the same increase.

Normally, in situations such as this, the principle investigator will either make corrections or do a retraction of the paper. This is of course an embarrassing, but an acceptable practice. In this instance, however, the principle investigator chose to do nothing; not a retraction, not a correction, not to have the original researcher repeat the experiment with supervision in case of error on their part, not to have other researchers in the lab continue to try with supervision in case of error on their part, or to repeat it in tandem with the original researcher to find out where the discrepanies arise.

Now, knowing what I know, I don't know what the ethical thing to do is. On one hand, it's none of my business, as I'm no longer in the lab and was never involved in this project. On the other, this data is now accepted by many in the field and quite a few people are spending an awful lot of time and money trying to obtain similar results for other products based on the published results.

In his favor, I do believe the principle investigator, and even the original researcher, believed their results to be valid. I do believe both of them to have been involved in very different but very unethical behaviour in the past however.

Anyway, the choices seem to be:
1) notify the schools board of ethics (risky as while I'm in a different lab, I'm still there).
2) notify the journal and hope to maintain anonymity.
3) nothing, just wait 'till other researchers are unable to replicate the results and question the paper's validity.

Thoughts anyone?

[Latrinsorm]

If the other researchers told you those findings in confidence, you can't do anything.

If not, I'd say go straight to the ethics board thingie. I don't know the ins and outs of scientific ethics, but it seems like the principle investigator should say SOMETHING if what ey's published has been rather convincingly disproven. The advantage to this is that if it turns out the principle investigator HAS done something or the other researchers were mistaken, you've kept it in-house but still made a reasonable request to the ethics board.

[Stanley Burrell]

Negative data is still very much data.

What field of science was this experiment in?

[SayGoodbye]

Why was it even published if it was known that the experiment couldn't be replicated?

Depending on what the experiment was, the results could be misleading and cost other researchers time, money and resources.

In the same situation I would anonymously (if possible) make the school's ethics board aware of the problem. But either way I wouldn't worry too much about it as you said, other researchers will find fault in the experiment and then they'll be looking for answers.