Twenty percent of scientists are crooks. At university in both Introduction to Criminology and the Senior Seminar in Criminolgy, our professor, Liqun Cao liked to cite an easy claim of “twenty percent.” The exact numbers – 18.3% or 23.9% – did not matter and neither did the specific study or survey. The teaching point was that criminality is not unusual. Heinous crimes are rare. Daily harms are all too frequent – and we all engage in one or another of them whether speeding or padding expense accounts. So, too, in science is it important to realize that fraud and misconduct in research are not rare.
False Prophets: Fraud and Error in Science and Medicine by Alexander Kohn (Oxford: Basil Blackwell: 1986), is a classic work that remains important. It sets a baseline for understanding fraud and misconduct in research. Kohn repeats famous cases such as Margaret Mead, Robert Millikan, and Trofim D. Lysenko. He also tells of N-rays, the Allison Effect, the Davis and Barnes Effect,, and polywater. From there, Kohn focuses on his special interest, clinical research. The book closes with chapters on broad and deep issues in ethics and science.
Marshall Thomsen of Eastern Michigan University has been teaching “Ethical Issues in Physics” for over twenty years. A search of “ethics physics” and similar items will return citations to Dr. Thomsen’s work at websites from the Illinois Institute of Technology, the University of Illinois, the University of Massachusetts, and Physics Today, among many more. When I had the class, he was on sabbatical and our professor was Patrick Koehn. The class also has been taught by Prof. Mary Elizabeth Kubitskey. Dr. Kubitskey’s master’s thesis was Teaching Ethics in a High School Physics Class.
The Office of Research Integrity of the US Department of Health and Human Services investigates and acts on cases of fraud in research when federal grant money is involved.
Friday, January 4, 2013
“Crime knows no neighborhood” is an axiom of criminology. In other words, every population contains members who stray from folkways, violate norms, harm others, break laws, ignore contracts, and betray trusts. Occupation, avocation, ethnicity, nationality, language group, religion, philosophy, ideology, age, sex, gender, height, weight, body mass, and shoe size are all irrelevant.
So, of course, some scientists are criminals. They falsify data; and they embezzle research funds. They also harass coworkers and subordinates, discriminate on the basis of race, age, religion, and gender. And they cheat on their spouses, beat their dogs, and kick their cats. But not every scientist who falsifies data abuses their aged parents. In fact, very few do. The arithmetic of intersecting sets limits the count. If 20% of scientists publish phony findings and if 20% of researchers carry non-existent students on their payrolls, then only 4% of research scientists do both.
At the same time, criminality is a way of life. The criminal researcher does not round up the value of a single point on one graph to make the curve smoother. And the vagary is not the first lapse after 35 years of devotion to truth. If a complete and nearly omniscient investigation could be conducted, it would most likely show falsified lab reports in ninth grade biology.
Of course, “most likely” is not “certainly.” When the case of Jan Hendrick Schön was finally resolved, the University of Konstanz revoked his doctorate, even though his dissertation was above reproach.
It seems that in the instance of Jan Hendrick Schön the pressure for results was his motive for crime. The pressure for results has been cited as a cause of research fraud. However, it is also true that truck drivers also labor under a call for results and that does not justify crime for them. Basically, everyone whether in a market economy or a centralized state is called upon to produce. At the end of a sabbatical, a professor is expected to show more than a sun tan.
Given all of the above, the research enterprise that does not engage independent investigation jeopardizes its funding and its social status. Which loss would be the worse is hard to say.Whether all crimes are evenly distributed across all neighborhoods is another question. It remains an easy assumption that life sciences are more susceptible to deviance than physical sciences. Tons of public money are thrown at both; but as living entities are more complicated than subatomic particles, experimental results may be harder to quantify rigorously. Confirmation bias may be a greater danger when we want to believe that we are helping other people live longer and better. Another explanation is that the US Department of Health and Human Services actually has an active Office of Research Integrity, while the U.S. Department of Energy has none.
On 6 April , a federal district judge in Boston, Massachusetts, dismissed a lawsuit that I had filed in 2009 under the US Freedom of Information Act. He concluded that the US government does not have to release a report on an investigation into a case of alleged scientific misconduct at a national laboratory. The ruling was disappointing but liberating: I finally had occasion to write about a case that has shown how the US Department of Energy (DOE) takes a strikingly hands-off approach to the oversight of such investigations.
“Misconduct oversight at the DOE: Investigation closed” by Eugenie Samuel Reich Nature 475, 20-22 (2011) Published online 6 July 2011 here
University oversight committees focus on human factors. We seek to protect individuals from unintended harm during experiments and surveys in psychology and sociology. But you cannot hurt a chemical or a star. Short of serendipity, we only find what we seek.
Previously on Necessary Facts