I have been writing about the problems of misconduct in research with academics having to resign as a result of misconduct.
Now comes a case of a Japanese anesthesiologist Yoshitake Fujii who apparently fabricated data on an epic scale for over two decades. About half of his incredible output of nearly 200 papers over this period (itself grounds for suspicion) are likely to be retracted by journals.
What is most disturbing is a new study published in the Proceedings of the National Academy of Sciences that says that misconduct lies behind most of the retractions, not error.
The field of psychology has been especially getting a black eye. Ed Yong, who has been doing excellent work covering this issue, says that Nobel prize winning psychologist Daniel Kahneman, whose work I have also been highlighting, has written to colleagues telling them that their credibility is in danger, especially in the area of priming research, and urges the adoption of an open and transparent replication process to prevent further abuses.
Scientists and academics have no real financial or political power. Any influence they possess comes from the faith that the general public has in them that they are knowledgeable and unbiased seekers of the truth. This image has been subjected to assault by those with overt political agendas because they do not like the scientific consensus in areas like climate change or health or evolution. They will use misconduct in one area to discredit all of academic research.
We should not give those people any ammunition. This requires academics to be ever more vigilant and put in place internal controls to minimize the danger of fraud and other forms of misconduct.
Clinical trials seem like a pain in the ass. In easy fields like chemistry and physics no one would think it is suspicious if you wrote 200 papers over 20 years.
I agree. There are far too many things being published in reputable magazines that are questionable and really don’t get sufficient scrutiny.
Unfortunately, you mention health, but one of the worst disciplines for questionable methods are the dietary studies. The vast majority of them rely completely on statistical evidence which, instead of leading to further causal research, are the final end product of their studies. There have been a few doctors that have pointed out the issues in this area of scientific study, but they are drown out by industries that rely on those conclusions to sell questionable products. One simple example, you still see most doctors recommending omega-3 fatty acids when the British Medical Journal debunked the benefits of them back in 2006 after looking through all of the studies (http://www.bmj.com/content/332/7544/752). And that is far from an isolated case; after looking through many dietary studies myself, I have found that the term “relative risk” (usually distilled to just “risk” in mass media) is used freely with any statistical variance even when statistical significance is not established correctly (or only barely reaches statistical significance).
In other scientific areas, I think honest scientists are facing issues with dishonest scientists. And the problem is that the general public doesn’t really have a method to distinguish between the honest and the dishonest. Until the dishonest get shoved out of the scientific community, I think this will be a constant struggle.
Jared A @1:
Please be kind to your host, and I’m glad you didn’t mention mathematics.
It can be easy for hard-scientists whose work is not tied closely to experiment to underestimate the difficulties faced by colleagues who are dependent on clinical trials, on interviews, and frequently on a host of people gathering data who are not under the principal investigator’s direct control. I had an unhappy foot-in-mouth experience (committee looking at a psychologist’s work) that forcefully demonstrated this fact to me.
The sad story of David Baltimore illustrates that having formal procedures in place and having an honest p.i. may not be enough to keep everybody honest.