Ethical Tremors in World of Science : Professional community alarmed by reports of dishonesty
Nothing is quite so corrosive of science as reports of fraud and misconduct by scientists. One expects cheating and thievery in business, where the object is to make money, but not in science, where the ultimate reward, ostensibly, is finding the truth.
But there have been so many allegations of scientific dishonesty lately that the British journal Nature observed that “outsiders can be forgiven for believing that science is not the honourable profession it pretends to be, but a mafia in which rival gangs scheme to win the credit for intellectual innovation.” The reports have damaged the credibility of science in Congress, which recently set up a special commission to examine the integrity of research supported by the U.S. Public Health Service.
The time has come for the scientific community to drop its defensiveness on this issue. It must help determine whether misdeeds are isolated events or signs of a wider problem driven by underlying institutional pressures to publish, succeed and win research grants from the government in a highly competitive environment.
CANCER-TEST SCANDALS: Leaders of science assert that fraud is rare and that, in any case, the process is self-correcting because other scientists ultimately will discover the error when trying to replicate the work. But the steady flow of reports of alleged cheating is disconcerting. Earlier this year the National Institutes of Health were jolted by reports that some researchers had compromised a breast cancer study by including ineligible subjects. More recently, the agency has been looking into whether data was altered in a study of ocular cancer.
Scientific misconduct ranges from outright fabrication and plagiarism and the sabotage of others’ work to the more ethically ambiguous acts of overinterpreting data to claim priority for an unproven idea or not giving sufficient credit to others. One hoary practice that should be dropped immediately is the use of “honorary” authorship by which a senior researcher puts his name on publications of disciples or co-workers even though he had nothing to do with the research and cannot defend it.
We suspect that instances of gross fraud are rare, if only because such cases are so easy to detect. When physicists at the University of Utah falsely claimed to have achieved nuclear fusion at room temperature in 1989, the conclusion was quickly debunked because it was so improbable. Probably the wider problem is that of fudging and shaving, the seemingly innocent corner-cutting that could pollute the scientific literature for years and is very hard to detect.
The problem is that there is little hard data on the true prevalence of misconduct. Five years ago Dr. Drummond Rennie, a professor of medicine at UC San Francisco, stirred a storm by proposing an experiment by which publications on clinical research would be checked to see whether they conformed to the raw data.
PUSH FOR SPOT CHECKS: That idea has been taken up by some members of the new Commission on Research Integrity, and is expected to come up at the commission’s meeting in Washington Dec. 1 and 2. Under it, journals that publish original clinical research would require that scientists submit to random, confidential spot checks of their records. Such audits of drug research by the Food and Drug Administration suggest that the incidence of misrepresentation and sloppiness is not small.
Outright deception in applying for or using federal grant money is grounds for criminal prosecution. But federal agencies cannot police the scientific community. That responsibility lies mainly with universities and hospitals. Staffs must be monitored better and institutions must examine themselves, asking whether they have put such emphasis on publication as a condition for tenure and promotions that the very scientific truths they seek to advance are compromised.
