Failure of Peer Review

[Papewaio]

I was channel surfing last night on TV and came across a nanotech show. They were talking about a physicist who was producing on average 8 research papers per day. Then I continued surfing on.

Today I was thinking that the show must have been an April Fools style show that was being shown out of season here in Australia. No way could a scientist continously produce a research paper every 8 days. So I googled Jan Hendrik Schön and the Wiki entry showned that he was indeed found to be a fraud.

Wiki Jan_Hendrik_Schon (http://en.wikipedia.org/wiki/Jan_Hendrik_Schon)

Then I found the transcript to the show itself and then last quarter of it was about his faking data. He was too lazy to even make up new graphs and just pasted the same data into multiple papers.

The Dark Secret of Hendrik Schön - transcript (http://www.bbc.co.uk/science/horizon/2004/hendrikshontrans.shtml)

What I found amazing that it took so long for people to smell a rat. It also shows a failure in the peer review process... timidity, lazyness, lack of proper dissecting of papers. The reason he got away with it for so long is that in the field of material sciences at nano scales it can be very difficult to reproduce materials and hence judge if what someone says is true.

Physics Web on Jan Hendrik Schön (http://physicsweb.org/articles/world/15/11/2)


In this case, the system worked. Science is self-correcting, as it is supposed to be. But we must not be complacent. If this kind of misconduct were to become commonplace, science would cease to be self-correcting and would be no better than any other belief system. Rooting out scientific misconduct in a sensible way will always be a grave responsibility for all of us.

The alarm bells sounded when Jan claimed that he had created a single atom scale transistor, this would have been a massive breakthrough worthy of a Noble Prize. At that point his papers came under far more peer review and it is at this point he was caught out.

[solypsist]

:burnout:
he became a victim of his own [self invented] success

[Major Robert Dump]

He should go to Korea and do some cloning research

[English assassin]

Well, he WAS cloning his graphs...

As a general observation, it is pretty difficult to detect fraud, at least at first and at least if the person committing it isn't completely stupid. (Like, say, announcing a nobel prize worthy breakthrough that would change the face of electronics) And would we want to? After all, the additional costs of a system where everything was checked not merely against the chance it was mistaken, but against the chance it was deliberately falsified, would be huge.

And does it always matter? IIRC its now generally accepted that Mendel fiddled his data, but he still got the right theory from it. And I would NEVER have finished my ****** practical courses at university if I hadn't often [rest of sentence deleted on legal advice]

[master of the puppets]

i wonder how rich he got off this lie, and i wonder what would have happened if someone in oppenheimers crew did that.

[Geoffrey S]

I like to think that if someone intends to use such a person's 'work' they'd examine for themselves the evidence and conclusions. The fact that the man faked his evidence was bad, but ultimately should not cause problems since if it reaches the point where someone may examine it for serious reasons it'd fall apart; only ridiculous levels of trust in the idea that he wouldn't fake such paper could cause trouble, but in such a competitive world of science I find this unlikely.

[BDC]

i wonder how rich he got off this lie, and i wonder what would have happened if someone in oppenheimers crew did that.
Probably not very. You can only sell stuff that works...

[Papewaio]

On a similar tangent:

900 university cheats busted (http://www.news.com.au/story/0,10117,17778634-421,00.html)

[rory_20_uk]

Those in the media are neither the brightest lightlbulbs, are not used to reading research papers nor are trained to do so and frankly have no interest in the truth if that ups the cost of the programme.

Fast, lots of graphics - forget the second you've watched it are the rules of modern TV.

~:smoking:

[doc_bean]

2 years ago, an ex-student body president (well, rough translation of the concept) graduated from my faculty with some nice results. The next year a doctoral student continue working on the subject but for some reason his test gave *strange results*, at least compared to the ones in the thesis. So after about half a year of trying to figure out what he was doing wrong (noone likes to admit they can't do something) he calls the graduate and asks about his thesis.
His reply:
" I'm graduated and don't have anything to do with the faculty or university anymore."

This might be an extreme example, but I've heard from a lot of people they 'adjusted' their data a little to make it fit better. I've even had assistants (doctoral students) tell me I should maybe, adjust a measured value a little bit. Personally i don't understand their problem, no one should expect perfect results, and mistakes in measuring do happen, you remove the points and mention it, no big deal normally, but it seems that in today's scientific world, honesty doesn't count for a lot anymore.

[Alexanderofmacedon]

In my 5th grade science project I tweaked a couple of grades...:shame:

I was young and I never did it again!:sweatdrop:

[71-hour Ahmed]

With the shear amount of papers and data out there I don't believe that anyone can feasibly tell from the limited data in papers if something is genuine or not if it looks right. And with so many journals it would be easy to replicate the same paper across several in different formats etc. At least one bunch of people in a Japanese Uni have put papers out on a subject closely related to my old research project that use a chemical that is fundamentally unsuited to the role its applied in. I know from my work that the results it gives are erratic (even with the controlling variable not altered, the measured variable shows a wide range of results, never mind what happens once you changed the control variable), yet there graphs are all nice and orderly with few results off the trend line and a very even trend. Mine in comparison would look like a picture of snow flakes quite easily.

[Zalmoxis]

So sad...