next »

[ssclark]

Purdue U. Reverses Course and Declares a Scientist Guilty of Misconduct

So, competitors cannot replicate the Purdue Prof's research and accuse him of faking his research results. Two panels address this and neither fault him on his research.  Rather, they fault him for comparatively minor (important, but minor) infractions that seem unrelated to the original accusation.  And these infractions were only found after close scrutiny of the Prof's day-to-day activities.

Put all profs to the same stringent test and you will find similar minor problems of gift authorships, hyperbole, etc.  Put anyone in any profession to such a stringent test and you will find similar minor problems.  To be sure, people need to be more transparent in their professional lives, but the infractions the Purdue Prof was found guilty of were not the basis for the complaint but found during the investigation of the complaint.

Committee no.1 finds no fault.  Committee no. 2 finds a little bit.  Which is it?  What was Purdue doing--trying the guy until they could find something to pin on him?  How would they have gone to do this?

In the professional world, the appearance of proper behavior is as important as behaving properly.  We need to have people believe in our research and have confidence in our findings.  However, institutions also need to appear to behave properly.  There is a case to made against Purdue that it went witch hunting in order to find or manufacture ANYTHING against this prof.  Even if this is not true, the perception of the University's behavior in this case is very troubling.

What a terrible blow for due process.  This is double jeopardy.  Purdue's faculty are the big losers here.

Steven S. Clark, PhD

[pyshnov]

ssclark:

So, competitors cannot replicate the Purdue Prof's research and accuse him of faking his research results.

Does it mean they are unable to replicate the conditions of his experiment? That's so strange. I remember that for some time the experiments with hybridization of DNA were considered irreproducible (that is, before the authors distributed important additions to the previously published technique).

However, I am told that there is no cold fusion. Is this not settled yet?

[ssclark]

This isn't exactly the cold fusion that was reported and unreplicated several years ago. 

It certainly is a problem when an experiment cannot be replicated and researchers can, and often do, go to great lengths to make sure that their experimental results, especially for seminal results, can be replicated prior to publication.  That is a good researcher.  In the Purdue case, it seems that prof was too sanguine about his research and did not do sufficient work to ascertain that the data were correct.  He should be sanctioned for that.

But, the point that no one seems to be talking about is Purdue's conduct in this case and that was the point of my earlier comment.  The appearance of impropriety on their part is very high.  Hopefully, this will not happen again.

Steven S. Clark, PhD

[alexking]

Whatever the pluses or minuses of this case and its outcome, there are some serious lessons to be learned, as if we did not already know them:

1.     "Failure to replicate" is never sufficient proof of a charge of scientific malfeasance.  Purdue erred quite seriously by pursuing this line in its original inquiries.

2.     Even if they don't get you for the original crime, engaging in a cover-up puts you at much greater risk of being found guilty of something. Taleyarkhan apparently erred in this regard, at least.  Cover-ups are always easier to demonstrate, and they will typically be more damaging than the original charge, too. Plus, their discovery is usually taken to justify the doubts about the integrity of the original work, rightly or wrongly, and without any of the hard work of proving them.

And perhaps less directly from the present case, but still of relevance:

3.     Honest errors are and always will be a part of the scientific process.  Let he or she who has never published an error cast the first stone. 

4.     The exposure of an error (be it deliberate or accidental) is a demonstration that the scientific process is still working properly, NOT that it is broken. 

Still, it is very disappointing to the scientific community when cases like this occur, and even more disappointing when they are handled poorly.

[jackit]

Note that this is just a reprint of one of the OP's blog entries.

Warts.

[daniel_von_flanagan]

Taleyarkhan is now appealing this decision, and also suing the scientists who have questioned his work (including colleagues at Purdue who claim he kept them from publishing conflicting results).  Not exactly normal scientific praxis.  - DvF

[santommaso]

It is not clear to me why the juridical concept of "double jeopardy" would be operative in a university setting. Universities can investigate as many times as they want to. The importation of legal terminology into a discussion of university policy may be rhetorically interesting, but it isn't helpful.