Scientific Fraud?

[Ringer]

I was reading an article in Retraction Watch about the rise in paper retraction. I was wondering two things really:

 

 

A.) Do you believe this to be because of stronger self policing, more people lying, etc.

 

http://pmretract.heroku.com/byyear ; a nice graph illustrating the trend from the site

 

B.) Do you think that publications are doing all they should be to make sure their readers know about these retractions.

 

 

. . . it’s somewhat difficult to know for sure, as Steen acknowledges, because

 

8% of retractions were for unstated reasons and up to 18% of retractions were for ambiguous reasons.

 

That lack of transparency resonated with us, as did another data point that Richard van Noorden, over atNature‘s The Great Beyond blog, picked up on:

 

Nearly a third of retracted papers remain at journal websites and do not have their withdrawn status flagged…

 

Specifically, 31.8% of papers were neither watermarked nor marked as retracted at their abstract pages. We’ve called for various forms of transparency before, including press-releasing retractions and writing clearer notices. But marking retracted papers to begin with is also necessary, and we’re disappointed to see journals doing this badly.

 

I think it's good more retractions are happening, if only because it shows an attempt to keep people honest. But at the same time the ambiguity with which some of the retractions happen is troubling. If mistakes were made I believe these mistakes should be openly revealed so others don't make those mistakes. Not only that, but keeping those withdrawn papers on websites and not immediately notifying subscribers about the retraction is disturbing.

 

 

[Mr Skeptic]

Why would this have anything to do with fraud? Or did you think scientists never make any mistakes?

[Ringer]

That's part of my point of ambiguity of some of the retractions, I never said there was anything wrong with retraction. The title of the thread was pretty much from the website. The issue I have is the reasons behind the retractions, whether it is because of increasing mistakes, increasing fraud, etc. That was my point, I said nothing about the infallibility of science, of which there isn't such a thing.

 

Do you not think that subscribers to journals should be made aware of these papers and the cause of retraction? If it was, in fact, a clerical error or something more dubious I would like to know. Also, it points out that even some of the articles that have been retracted are still being cited, I can't remember the exact paper they mention, and about half was actually using it as a reliable source.

 

[edit] I realize that I meant to put a question mark in the title. Failing to do this may make this thread seem more like a statement saying that there is, in fact, fraud happening. I meant this to be framed in the form of a question, if this is the cause of misunderstanding I apologize.[/edit]

 

!

Moderator Note

Fixed

[lemur]

This is an interesting topic. It would be interesting to survey the reasons for retraction to see if there are patterns. It makes you wonder if there are some cultural changes going on in academic soul-searching, or if it's related to shifting knowledge and/or analytical consciousness. Or maybe it is just coincidence or a trend of increasing doubt due to persistent threats of budget-cuts and/or other personal, professional, and/or economic insecurities.

Edited January 18, 2011 by lemur

[CharonY]

Well another aspect may be that the competition pressure between scientists is ever increasing. There is an ongoing trend of having more publications and in some cases this will lead to overhasty publications that may be erroneous. Of course, this climate may in some cases may encourage fraud (especially if the choice is between tweaking a "minor" data set or losing your job). However, I would say that outright fraud is still very rare.

Publication of crap, on the other hand, is not that rare.

[ajb]

Do you not think that subscribers to journals should be made aware of these papers and the cause of retraction? If it was, in fact, a clerical error or something more dubious I would like to know. Also, it points out that even some of the articles that have been retracted are still being cited, I can't remember the exact paper they mention, and about half was actually using it as a reliable source.

 

Making people aware of the reasons could be a good idea. The authors themselves could state why as part of the retraction.

 

I don't think many retractions will be due to clerical or editing errors. Typos and similar always seem to creep though. I have yet to totally misunderstand a paper because of that.

 

For minor errors or some oversight a correction can often get published. So, more minor problems can be fixed.

 

I don't think there is anything wrong with citing papers that have clear errors in them, as long as who is citing them knows this. I mean, part of the work may well still be correct and interesting. Also, understanding why it is wrong can be very illuminating.

[Ringer]

I agree that knowing the errors, if they're not just typos and such, could be illuminating to future research. But the problem is not being aware that there could be errors in the papers, not even on the online journals. 18% retractions for ambiguous reasons and up to 8% for unknown reasons. I don't like the idea that I read and cite a paper that was retracted without me being aware or, if I am aware, without knowledge of what the problem with the paper was. It could be because of a typo that has nothing to do with the results, yet I no longer want to use the paper because I'm no longer sure how reliable it is.

[JohnB]

Fraud is possible, but I think quite rare. Fraud being the doing of something with the intent to mislead.

 

Confirmation bias and the pressure to publish along with "publication bias" may have a lot more to do with the number of retractions. Reading Retraction Watch recently I was rather astounded by the Editors response to the question "Why was the paper retracted?"

 

It’s none of your damn business.

 

I would have thought that the reason for the retraction was everybodys business. Letting other researchers know where the mistake was will save them time and effort.

 

Retraction Watch article here.

 

There was also an interesting piece in The Guardian on this topic.

 

In the case of one retraction the reason was;

"inaccurate DNA hybridization detection results caused by application of an incorrect data processing method"

 

Which is not particularly informative. Retraction Watch got in touch with the authors and recieved this reponse;

he explained that they forgot to correct for something in their analysis, which made the technique they were testing appear to be more powerful than it really was, and actually they found it's no better than the original process it was proposed to replace.

 

No fraud, just a mistake that led to a wrong conclusion. When the mistake was realised, they retracted the paper. No big deal.

 

Also of interest and perhaps relevent to this topic is this rather longer piece in the New Yorker on the "Decline Effect" and the difficulty some scientists have in replicating their own experiments. Strong findings early on that diminish with time may lead to retractions, or the original papers simply fade away and are no longer referenced.