The Astrophysics Spectator

Home

Topics

Interactive Pages

Commentary

Other Pages

Information

Commentary

The Unethical Advantage

One of my forays into data analysis is a study of whether gamma-ray bursts are clustered in time. It is a simple paper that shows gamma-ray bursts were uncorrelated in time, meaning that each gamma-ray burst appear to be the sole event produced by its source. As I placed the finishing touches on this paper, a visiting assistant professor strolled into my office and sat in a chair next to the desk. “Put my name on your paper,” he hissed. I was stunned. I had known this guy for many years, and I knew he would try to weasel his way onto any paper he heard of, but I never expected such a forward demand. “If you don't put my name on your paper, I'll scoop you!” I knew lack of time was the reason he worked so hard to get listed onto papers he had no role in, so his threat was a pure bluff. I made some noncommittal comments to get the guy out of the office, ignored his demand, and from that point forward I made certain never to discuss my work within this fellow's earshot. I also complained to the more senior members of the group, including my supervisor.

This little event, one of several examples of unethical behavior I have seen in science, came to mind last week when I came across the recent commentary by Martinson, Anderson, and de Vries in Nature of their research on academic dishonesty.1 These authors sent surveys to U.S. researchers receiving grants from the National Institute of Health; the researchers were asked whether they had engaged in any of the conducted listed in the survey over the past three years. The surveys are not linked to the names of the researchers to keep the answers confidential. The mailing resulted in 1768 useable returned surveys out of 3409 delivered surveys.

Some of the unethical activities listed in the survey, such as “ignoring major aspects of human-subject requirements,” are specific to medical research, while others are more about sloppy work than unethical behavior, but the remaining activities occur across all branches of science. These universal acts of unethical behavior, numbered as they are in the survey and followed by the percentage of scientists who say they engaged in the activity, are

  1. Falsifying or ‘cooking’ research data—0.3%.
  2. Using another's ideas without obtaining permission or giving due credit—1.4%.
  3. Failing to present data that contradict one's own previous research—6.0%.
  4. Overlooking others' use of flawed data or questionable interpretation of results—12.5%.
  5. Publishing the same data or results in two or more publications—4.7%.
  6. Inappropriately assigning authorship credit—10.0%.
  7. Dropping observations or data points from analyses based on a gut feeling that they were inaccurate.—15.3%.

The authors say that the university compliance officers they spoke with consider the first-ten items on their list as actions that would result in disciplinary action.

The survey shows only a small number of people admitting to the most serious actions. I don't find this a surprise for the first item, because you simply cannot help yourself by falsifying data. If you invent numbers, other researchers will simply show your numbers to be wrong, so you simply portray yourself as incompetent. As far as publishing the ideas of others as your own, I have seen this occur often enough in astronomy and astrophysics that I believe it is much more common than portrayed by the survey.

Where the survey is most interesting, and this is a point hammered home by the authors, is in the relatively large number of researchers engaged in less notorious acts of dishonesty. The authors describe these actions as corrosive to the integrity of scientific research.

At first glance acts such as publishing the same results in multiple research paper and inflating an author list on a paper to include people unconnected with the research appear to be harmless. In fact, they are quite destructive to the quality of research. As sports is plagued by performance-enhancing drugs, so the research community is plagued by bibliography inflation. Research papers are critical for obtaining a job, landing research funding, and winning tenure. You want your name on as many research papers as possible. This means that the rewards of cheating the system by inflating your bibliography are very high, and if ethical standards are not enforced, this cheating in a tight job market gives the advantage to the dishonest. In sports, we do not end up with the most gifted and hard-working athlete, but with the athlete most willing to drug himself; in science, we end up not with the best researcher, but with the researcher most willing to act unethically.

If I were to survey astronomy and astrophysics about unethical actions, I would add three more actions to this list.

First, have you every treated publicly-funded data as proprietary data? I have seen some researchers treat data as private property, even when the data is collected from a publicly-funded machine or satellite. There is a good argument to be made that the researcher who conducts an observation should have sole access for a limited time, but after a fixed time data should become public. NASA to its credit has and enforces such rules.

Second, have you used the referee process to suppress the articles of competitors and promote the articles of colleagues? I recall a colleague bragging that he had rejected a paper of someone he had taken a dislike to, I have seen papers held-up for trivial reasons, I have seen papers with serious flaws accepted within a week of submission, and I have seen referees wait many months before submitting a referee's report that demands substantial revision to the paper.

Finally, have you turning a blind eye to the scientific misconduct of a colleague. When I complained about the visiting assistant professor who tried to force his way on my paper to one of the senior scientists in my group, his reaction was “well, the guy is acting crazy because of the stress of getting tenure. Once he gets tenure, he'll straighten out.” Now, this fellow did get tenure, but I doubt he has straightened out. After all, he paid no penalty for his dishonesty.

Martinson et al. place their emphasis on structural changes to lessen the incentive for unethical conduct. Their main concern is with perceptions of the unequal distribution of resources among researcher. I think a bigger problem is that in an extremely tight job market, where the likelihood is you will be changing careers before many years have passed, there is effectively no penalty for unethical behavior, but tremendous benefit. Unethical behavior will continue to flourish until this behavior carries much greater risk than reward.

1 Martinson, Brian C., Anderson, Melissa S., and de Vries, Raymond. Scientists Behaving Badly. In Nature, vol. 435, p. 737-738 (June 9, 2005)

Jim Brainerd

Ad image for The Astrophysics Spectator.