Mar 30 2015

Peer-Review Scandal

Biomed Central, a UK company that publishes 277 peer-reviewed journals, announced that it is retracting 43 articles because of “fabricated” peer-review.

Peer-review is a process that many scientific journals use to vet submitted articles. Typically an editor will review the article but also send it out to two or three experts in the subject matter and have them take a close look at the article to make sure everything is high quality. Most submitted articles will come back with required changes before acceptance. Of course many articles are rejected outright.

The process is not perfect, but it is one critical layer of quality control. The “peer-reviewed literature” is therefore a body of evaluated knowledge that has met at least a minimal standard of quality.

Of course, your mileage may vary. Not all peer-reviewed journals are as rigorous. Also, whenever there is any system in place to separate the wheat from the chaff, someone will try to game the system for their own advantage. There also needs to be some monitoring or policing in place to ensure the integrity of the system.

Scientific fraud is a terrible thing. The institution of science requires complete transparency, and fraud violates transparency and reduces confidence in the whole system. Peer-review fraud is not as bad as fabricating data (the ultimate scientific sin) but it’s bad.

The concern is that the problem may be much deeper than these 43 retracted articles. This story, therefore, may not be over yet. I don’t think this will turn out to be a systemic problem but it may be widespread.

The Washington Post reports:

A partial list of the retracted articles suggests most of them were written by scholars at universities in China, including China Medical University, Sichuan University, Shandong University and Jiaotong University Medical School. But Jigisha Patel, associate editorial director for research integrity at BioMed Central, said it’s not “a China problem. We get a lot of robust research of China. We see this as a broader problem of how scientists are judged.”

That’s interesting – it seems that the problem, while not limited to China, may be concentrated there.  This brings up another highly sensitive issue. Science is both an institution and a culture. While there certainly is a universal culture of science, there are also subcultures – in different disciplines, and in different countries.

China’s scientific culture is embedded in China’s broader culture.

One potential problem arises when a particular scientific subculture includes an excessive deference to authority, without the proper balance of respect for freedom and openness. Junior researchers who feel they cannot challenge their superiors, or feel incredible pressure to get positive results, are more likely to bend or break the rules to please their boss. This can be poison in a scientific environment.

I will note, as I have in the past, that a review of the acupuncture literature, for example, showed that 100% of studies coming out of China are positive. This is not statistically possible, even if acupuncture worked, which I think, given the totality of evidence, it doesn’t. This is evidence of systematic bias – with 100% penetrance.

I am not trying to point the finger at one culprit, but rather just pointing out that ethical problems in science are not evenly distributed. This provides an opportunity to examine the institutional and cultural factors that may contribute to scientific misconduct. I agree with Patel that we should not overstate the significance of clumping in China or pretend the problem does not exist elsewhere, but neither should we ignore it. It’s data.


This latest scandal will hopefully provoke another round of self-reflection among scientific journals and improve quality control.  Quality control is an endless game, and there will always be problems like this. At least the peer-reviewed literature has the power of retraction, to purge itself of fraud or poor quality when it comes to light.

Like this post? Share it!

7 responses so far